The key landed with the streaming attachment ceiling without a class, so the
settings package panics at registration and every test importing it fails.
A stored reference carried a path the adapter opened as written, so a session
file that was edited, restored or imported could name a file no tool had
checked. read_video now records the path it verified under the file's hash in a
process-local registry; the adapter resolves through it and reports an
ungranted reference as video.file_unavailable.
Account caught with String(e), so a coded refusal such as
account.signin_disabled printed "Error: " and the kernel's English
fallback instead of the wording kernel.ts already has, and its
expired-request message was set without t() and had no English. The
port's accountPoll also hand-rolled its fetch and threw the raw
response text, a JSON envelope with no code to word.
Catch with reason(e) like every other panel, put the expiry message
through t() with an English entry, and send accountPoll through post0.
Problem: ordinary binary uploads can consume unbounded workspace disk, and the
review needs direct coverage of file versus directory grants at read consumers.
Root cause: the streaming saver copies until EOF without a byte policy; the
project-config merge must not become an authority for a user disk safeguard.
Fix: enforce a named 1 GiB default while counting real bytes, retain rooted
cleanup on overflow, and return typed ceiling/source facts. Allow user config
to set the limit or explicit zero, hold it across project merges, render it
only at user scope, and strictly read it without degraded config fallback.
Cover selected-file siblings and directory link escape through actual reference
and tool consumers. Incorporate current Studio without rewriting history.
Verification: the former 8-byte/9-byte acceptance fails first and now refuses
cleanly. Full config race and focused control/serve/builtin races, frontend
tests/build, actual Electron rejection/retry, native Windows ARM64 boundary
and cleanup regressions, pinned lint and repolint pass.
Problem: the attachment PR conflicts with Studio's extracted composer helpers
and the newer worktree setup surface.
Root cause: both branches changed the Composer's chip type and helper ownership.
Fix: merge origin/studio, use its shared composerChips owner and worktree setup
flow, and retain cancellable attachment uploads and native path references.
Remove the redundant local chip type instead of duplicating the shared owner.
Verification: TypeScript checking, 155 focused frontend tests, production build,
attachment/CSRF/cloud/control/fileutil Go regressions, 169 Electron unit and
packaging tests (one existing conditional skip), pinned lint and repolint pass.
Review follow-up: the audit helper used unix.Flock with no build tag, so lint's
typecheck and test (windows-latest) failed with 'undefined: unix.Flock'. Take the
lock through internal/base/filelock's TryAcquire, which has the same unix and
windows implementations the store uses, so the held-lock tests run on every platform.
saveNetwork, uploadWallpaper and clearWallpaper hand-rolled their fetch
and threw a plain Error, so the code the kernel sent never reached
reason(): an unsupported wallpaper read as the kernel's English fallback
in a Chinese window, a failed clear as a path and a status, and the
existing wording for network.rejected and wallpaper.* went unused. They
now go through post0 and del, whose fail() keeps it.
SaveNetworkSettings also stops prefixing its refusal with a Chinese
sentence: the frontend frames network.rejected itself, and the prefix
would read twice in one language and as a fragment in the other.
SsePort still declared exportPlugin, dryRunHook, trustMcp, reconnectMcp,
parseMcp and probeProvider after they were split into its faces, and a
method on the outer class hides the face's. So the faces' dryRunHook and
probeProvider, which go through post0 to keep a refusal's code, never
ran: probeProvider threw the raw response text, and onboarding showed a
refusal envelope as its message and could never offer the vendor-console
link it keys on provider.probe.unauthorized. exportPlugin was the other
way round, the face holding the stale copy.
Drop the six copies, move the current exportPlugin into SseExtensions,
and guard the chain so no face redeclares one below it.
renderSolveProfiles folded every run that reached a correct checkpoint into
the boundary medians, including runs whose trajectory was never read. With
-checkpoints and no -trajectories those runs carry zero round and
verification tallies, so the report printed "rounds before 0 · rounds after
0 · verifications after 0" as if they had been measured.
Only runs whose split was actually recorded now fold in, and a suite where no
run recorded one says the boundary was never split instead of printing zeros.
The all-measured medians are unchanged.
Problem: creating a worktree dropped keyboard focus when the composer became
inert, returning to editing did not restore it, and terminal pages still
reported waiting or mapped cleanup failure to a confirmation state.
Root cause: draft restoration attempted focus before React removed inert;
the preparation page had no focus owner, phase fallbacks conflated terminal
states, and details/edit actions shared a configuration identity.
Fix: focus the preparation heading on entry, recover focus when a removed
phase action leaves the body active, preserve surviving detail focus, and
restore the composer only after the stage-clearing commit. Use distinct
view action IDs and explicit terminal labels, and show elapsed waiting only
for active phases without changing existing typography or button recipes.
Verification: failure/cancellation interaction regressions, phase/disclosure
unit tests, real-browser focus and draft-retention checks, English dark-page
labels and shared button hierarchy, frontend typecheck/build/census, and
repository lint/check; full frontend: 634 files and 5,257 tests passed.
No Git or backend behavior changed.
setShellKillTree on Windows ran its own taskkill /F /T inside
exec.Cmd.Cancel with no time limit. #13323 bounded the taskkill in
proc.KillTree because Wait does not return until the cancel hook has,
but this hook never went through KillTree, so a PDF text extraction past
its 8 s deadline could still hang on a taskkill that never answers.
Delegate to proc.SetCancelKillsTree, the helper proc.RunCommand already
uses, so the only taskkill left is the bounded one in proc.
Review follow-up: Partial now accumulates across legs (a killed first leg plus a
complete second leg is still a lower bound), and the meter comparison skips a
Partial run the same way it skips an unaccounted one, since its missing tail is
not a divergence. Rebased onto current studio; the stub tests use requireShellStub
so test (windows-latest) can run.
Problem: hosted atomic coverage intermittently lost the MCP resource tool,
and CodeQL reported an unchecked filesystem probe during worktree cleanup.
Root cause: a queued cache callback read clients after Host.Close cleared
handshake metadata. Cleanup accepted a result path without independently
checking the trusted managed root used by its construction owner.
Fix: snapshot advertised capabilities and instructions before releasing the
deferred spawn; keep only persistence asynchronous. Capture the managed root
once for creation and rollback, validate a local relative descendant before
probing the filesystem, and retain registration and unchanged-head checks.
Verification: deterministic shutdown/cache and wrong-boundary regressions;
plugin atomic coverage and race tests; boot atomic coverage; worktree/serve
race checks; native Windows rollback cases; local CodeQL path query; root
lint, Windows lint and changed-file repository checks.
A leg whose final record was missing but whose <metrics>.partial snapshot was
read was marked Unaccounted, which gatherSuiteStats skips outright. That
contradicted Partial's meaning — real numbers that stop at the last snapshot
and belong in the totals — so the recovered tokens, cost and lower-bound hint
were all unreachable. Only a leg with neither record is now Unaccounted; a
recovered snapshot stays accounted and sets Partial, matching the semantics
swebench mode already uses.
renderMeterAccounting also treated PromptTokens+CompletionTokens == 0 as an
absent harness measurement, inferring state from a value. A persisted record
saying zero is a valid zero, so the guard is now the explicit Unaccounted flag
alone.
Tests: the copied audit case that asserted unaccounted=1 for a sidecar-recovered
run encoded the wrong semantics and now asserts Partial=true, Unaccounted=false
and that 1200/80 and the cost reach the aggregates; the divergence fixture that
relied on zero tokens sets Unaccounted explicitly, and a new case pins that a
persisted zero is priced on both sides. The record-less case and the healthy-run
control are unchanged in intent.
Problem: failed or cancelled worktrees remain registered on Windows, and the
cleanup-failed screen leaves users without a concrete recovery procedure.
Root cause: native backslash paths are compared directly with Git porcelain's
slash paths. Cleanup uncertainty deliberately fences replacement creation.
Fix: normalize both registered and expected paths while preserving symlink
resolution and the unchanged-head guard. Cover native rollback and changed
branches, wire the Windows regressions into PR CI, and explain draft retention,
manual reconciliation, runtime closure and retry from the source folder.
Restore the existing design QA content and metadata, append only workspace
acceptance, and remove all eight screenshots from the final PR tree.
Verification: native Windows regression fails before the fix; the fixed worktree
package passes 36 tests with 6 platform skips. macOS worktree suite, focused
serve/worktree race tests, Windows lint, workflow contracts, 5,245 frontend tests,
typecheck/build, census and repository lint pass. Hosted CI still needs approval.
Problem: Windows smoke reported an empty live catalog in the disabled-MCP
effect test's fresh always-loaded phase; the fixture expected one read tool.
Root cause: Build returns while cold discovery and always-loaded preconnection
can still be in flight. The fixture immediately issued an explicit connection,
leaving competing handshakes and initial catalog publication in its setup.
The adjacent always-load effect tests already await the boot-owned connection.
Fix: Wait for that connection in cold and always-loaded phases before explicit
registration. Keep cached deferred startup explicit and retain all filtering,
restart, catalog and provider assertions. Report held tools and server catalog
diagnostics if the allowed tool is still missing.
Verification: Related disabled/always-loaded/held-definition effect tests pass
normally and under the race detector. go vet, make lint and make check pass.
Full builtin and boot suites pass as well.
Native Windows CI must confirm the replacement candidate.
Problem: an explicit source refresh can repopulate the palette after switching
workspace, port or session, and an older initial read can overwrite a later read.
Root cause: only the initial fetch had a lifetime fence; manual reads published
without a shared generation. The browser guard still awaited a retired hover card.
Fix: put initial and manual reads behind one latest-read generation, invalidate
it on scope changes, and verify the unified palette before and after task delivery.
Verification: four deferred-response regressions fail without the fix and pass
with it; 633 frontend files / 5,245 tests pass, typecheck/build, census invariants
and fixture verification, repository checks, and browser palette interactions.
Upstream workflows remain pending repository-admin approval, not passed CI.
Problem: the PR overlaps the current Studio font-scaling update.
Root cause: both changes replace the workspace stylesheet, and hub additions
bring its combined source over the repository file-size ceiling.
Fix: retain the new palette with shared, scalable typography tokens; preserve
all upstream behavior and extract hub close/shutdown ownership together.
Verification: 633 frontend files and 5,241 tests; typecheck/build; affected
worktree, serve, control and boot suites and race tests; Desktop Go suite;
make lint/check; fresh production browser palette with no console errors.
Problem: workspace configuration disappeared after the first task and creating
an isolated worktree left the composer without visible progress or recovery.
Root cause: workspace selection, Git branch switching and runtime construction
had separate UI lifetimes, with no cancellable preparation/publication boundary.
Fix: keep one searchable workspace palette across conversation states; expose
real preparation phases with operation-ID reconciliation, retry and cancellation.
Serialize cancellation against publication, terminate checkout descendants and
clean unpublished resources before acknowledging cancellation. Preserve task
and attachment drafts, and reuse existing typography throughout status screens.
Verification: 5,231 frontend tests, typecheck and production build; root and
Desktop Go suites; affected backend race tests; lint and repository checks;
real Git browser checks for switch, failure/retry, cancellation and handoff.
Include light/dark and wide/narrow before-and-after screenshots and design QA.
Problem: CodeQL still reported a path-injection alert at the worktree
registration read, and a registration symlink could read outside its metadata
directory before assigning the checkout to a project.
Root cause: Validating checkout identity and the registration backlink did not
constrain the filesystem operation that reads that backlink. The plain
os.ReadFile call followed symlinks outside the selected Git directory.
Fix: Validate the checkout metadata path relative to the shared repository,
open that directory through os.Root, and read the fixed gitdir entry through a
second root confined to this checkout's metadata. Preserve relative backlinks
and internal symlinks while refusing escaping registrations.
Verification: A real-Git symlink-escape regression fails on the original code
and passes with the fix. Worktree/workspace identity suites, grouped-project
HTTP regressions, focused race tests, Windows test compilation, go vet,
make lint and make check pass. Local CodeQL CLI 2.27.1 with go-queries 1.6.11
reproduces the exact remote alert before the change; afterward that alert is
removed and the same path-injection query introduces no new results.
Builtin and boot integration suites also pass.
Problem: Checkout labels could push a conversation title beyond its breadcrumb
at phone widths and high interface scale.
Root cause: The fixed-width directory badge consumed the title's reserved
space even after project and folder controls were hidden by the narrow layout.
Fix: Group project and checkout context in a shrinkable clipped flex child,
allow labels to ellipsize, and preserve the title's existing four-em floor.
Document the shared layout constraint without weakening browser guards.
Verification: The eight failing light/dark width-and-zoom cases now pass in
perf/narrow.mjs. Title spacing and overflow, intrinsic layout and hover guards,
12 related interaction tests, typecheck, production build, vet and repository
checks pass. The Go implementation is unchanged by this layout correction.
Problem: Grouped projects could open the source after an explicit worktree
selection, hide source-alias history, omit inferred launcher exclusions, or
remain unusable when their source folder disappeared. Copied Git metadata
could also claim another checkout's project identity.
Root cause: Display identity was reused as execution and catalog authority.
Project inference duplicated an unbounded marker walk without validating
managed checkout containment or the reverse Git registration.
Fix: Preserve selected directories and all catalog members, choose a surviving
project directory, and use complete member roots for editor discovery and
launch. Reuse shared checkout identity and validate its managed boundary,
local prefix and registration backlink. Merge studio, move root resolution to
its project owner to preserve the file-size budget, and align the title guard
with the intended checkout label while retaining spacing and overflow limits.
Verification: Failing original-head HTTP/path regressions pass after the fix.
Frontend tests (629 files, 5,221 tests), production build, focused race tests,
go vet, make lint and make check pass. Added a 100-root tree benchmark and a
deterministic delayed-header test; repository suites and browser guards cover
adjacent behavior. Hosted checks must validate the replacement PR head.
TestCredentialSinkRegistry walks internal/frontend/serve, where serve's
own tests create and remove extra-* folders through tempAside. Run in
parallel with those tests, the walk can list such a folder and fail when
it is gone a moment later, the same race that failed the storage
state-path guard. Skip a vanished entry with testenv.VanishedMidWalk.
Problem: Studio lists a managed worktree and its source checkout as two
same-named sidebar projects, scattering one project's conversations.
Root cause: Sidebar ownership, session catalogs, and project operations use
checkout paths as project identities even when they share a Git repository.
Fix: Resolve project ownership from managed checkout Git identity while
retaining each transcript and runtime working directory. Merge session
catalogs before paging and keep registration, ordering, capacity, and removal
consistent with the project group. Show the checkout type in the chat header
and mark isolated history without migrating persisted data.
Verification: Serve/worktree/workspace identity and builtin/boot suites,
focused race tests, Desktop Go tests, frontend tests and production build,
go vet, make lint, and make check passed. Browser interactions and an installed
macOS package verified grouping and original/isolated conversation navigation.
GET /workspace/file takes an optional root, checked against the grant list and re-vetted on read through the decision the picture route uses. The workbench carries the folder, never saves, and says why.
A queued message that froze an image cannot be admitted as mid-turn guidance. The manifest now records that (images), and the queue row says so instead of 'not delivered yet'.
recordStalled emitted a notice with no code, so the frontend showed the kernel's English sentence as is. It now carries job_stalled and a JobNotice payload with the quiet span, like the other job notices; the English text stays the fallback.
Hosted Web enables the multiplexed event stream for every pane, but a
Studio kernel without GET /events/mux answers it with the SPA shell. The
mux client waits for the stream's open frame to name its connection, so
no pane ever subscribed and none fell back to its own stream: the page
kept loading and stopped receiving events until a reload rebuilt it from
a snapshot.
RemoteEventSource is a polled stand-in that cannot carry that protocol,
so a hub reached over the relay now reads each pane's own /events. A
local hub keeps the shared stream.
The event contract is unchanged: attach, resume, cursor, watermark and
the pending-prompt attach all keep the code paths PR #13499 fixed.
Problem: CodeQL reports uncontrolled paths in streamed upload cleanup and
attachment creation. A replaced storage directory can redirect failed-upload
cleanup to an external same-name file; a linked .reasonix parent can escape too.
Root cause: filename validation and path-based writes/removals do not preserve
directory identity across filesystem changes during I/O.
Fix: use os.Root for workspace-contained directory/file creation, share an owned
file lifecycle between streamed and legacy byte writes, and retain the directory
handle for failure cleanup. Publish the existing untracked marker atomically
through the same rooted directory rather than reopening its pathname.
Verification: full control/serve/fileutil suites, focused race tests, vet,
pinned linter and repolint passed. Native Windows 11 ARM64 regressions passed
with directory junctions and protected cleanup, plus large bounded streaming.
The independent replay test failure also reproduces with the PR's Go changes
removed through a base-source overlay; no replay behavior or CI gate is changed.
Fold the fragments merged after the first notes commit into 2.43.0.md and remove every consumed fragment, so the release commit leaves release-notes/studio/unreleased empty.
Problem: Review requested a user-visible release fragment and clearer memory
bounds. The live regression also assumed new-session activation must create
a second runtime, which is false when an existing blank pane is reused.
Root cause: A workspace may already own a reusable blank pane, particularly
when native Windows roots have matching canonical spelling. The old setup
waited for a POST and a new pane identity instead of editor readiness.
Fix: Wait for boot completion at a stable viewport, allow a blank-pane reuse,
and preserve assertions for the captured workspace, one shared intake route,
saved upload size and four ready attachments. Allow an explicit test host
binary. Add bilingual release notes and state that memory, not total disk
consumption, is bounded by streaming.
Verification: Windows 11 ARM64 control, stream, HTTP upload/image-limit and
exact-file read-tool regressions passed. Electron 44.2.0 exercised picker,
paste and drop of a 111 MiB file and a 65 MiB streamed File with four ready
attachments. Release-note tests, repository checks and script syntax passed.
Problem: The full frontend suite rejects the upload abort listener as an
unresolved interaction registration.
Root cause: The syntax-based census does not infer a Request.signal member's
AbortSignal type and cannot distinguish it from a possible user-input target.
Fix: Keep an explicitly typed upload signal and teach the shared receiver
certifier to recognize scoped AbortSignal bindings. Preserve unresolved
receivers and refuse borrowing an outer signal type across shadowing.
Do not add a file exception or raise any census or lint ceiling.
Verification: Abort receiver, actions census, cloud upload and attachment
regressions, frontend typecheck, production build and repository checks pass.
Problem: Large ordinary file attachments cause avoidable whole-file
allocations and exceed the legacy JSON attachment transport limits.
Root cause: Picker and paste intake Base64-encode files even when the shell
already knows their native paths. Pathless uploads also pass through JSON.
Fix: Share native path references across picker, paste and drop. Stream raw
ordinary-file bodies to bounded disk writes and clean up failed or cancelled
uploads. Grant external-file access only to the selected file. Preserve
image, legacy JSON and cloud relay limits; bound cloud intake before
materializing an oversized body. Document the transport and live regression.
Verification: Owning Go package suites, focused race tests, go vet, lint,
repository checks, frontend typecheck, production build and attachment
regressions passed on the original tested baseline. Real macOS Electron
exercised native picker, paste and drop with a 111 MiB file and an exact
65 MiB pathless upload. Windows native runtime verification remains open.
A window that gives each pane its own EventSource spends one HTTP/1.1 connection per pane, and a browser origin holds six, so the sixth stream starves every stop, submit and read behind it. GET /events/mux is one stream for all panes of a window; PUT/DELETE /events/mux/{conn}/panes/{id} subscribe a pane with its own cursor. /events and /events/replay are unchanged.
RemoteEventSource now leaves the attach cursor empty so the kernel hands a late client the pending prompt. The tree stream shares that class but reads the tree rather than history, so an empty attach for it re-read the tree on every poll and failed tree_events.test.ts. The attach query is now chosen once from the path: the main stream attaches with no cursor, the tree attaches past the watermark as it always did.
Verified with the whole frontend suite: typecheck, 5139 tests, census, census:verify, build, and no fixtures in the bundle.
Home declares no entries, so the survey treated it as sole occupant of its
directory and counted the sessions and archive that state declares in the
same folder (one folder on a default Windows install).
The cost chip drew a literal "¥" default, so a fresh session read
"本轮 ¥0.00" in a currency nothing had priced, whatever the display
currency was. Leave it out until the first priced round or restored total.
The polled attach took its pending-prompt snapshot under promptEmitMu but read the broadcaster watermark after that lock was released. A prompt raised in between was numbered at or below the watermark the client then adopted as its cursor, so the next resume skipped it — the original bug in a narrow window.
The watermark is now read in the same critical section as the snapshot, and before it: a prompt raised after that point is numbered above the cursor the client adopts, so the worst case is a prompt named twice rather than one lost. writeEventsReplay reads the watermark once and passes it in, instead of the writer reading it after the lock.
The shell's no-store comment is trimmed to the comment budget so the lint gate passes.
The argument completion of /goal, /memory, /mcp, /skills, /plugins, /theme,
/currency, /reasoning-language, /ask-timeout and /feedback returned English
literals; read them from the i18n catalog like /language and /effort do.
TestEveryTopLevelStatePathIsAnOwnedEntry walks every package under
internal/ and failed the moment WalkDir could not open an entry. The
serve, control and config tests take their extra folders from their own
package directory (tempAside), so when go test runs packages in
parallel the guard can list one of those folders and find it gone a
moment later:
open .../internal/frontend/serve/extra-3867033970: no such file or directory
The config package's state-root walk already tolerates this through
vanishedMidWalk, including the sharing-violation and access-denied
errors Windows gives for a directory being deleted. Move that helper to
testenv as VanishedMidWalk, unchanged, and use it from both walks.
TestCreateFromBranchKeepsSourceAndSeparatesTasks required the checked
out README to be exactly "base\n". Git for Windows ships core.autocrlf
true in its system config, so the worktree holds "base\r\n" and the
test failed on Windows CI. Trim it as the sibling dirty-source test in
worktree_test.go already does; the check still tells the selected
base apart from the current head and the dirty source.
TestAuditLegacyCacheUpgradeReportsOnlySuccess blocks the legacy-key
upgrade write by chmod-ing the cache dir to 0555. On Windows that only
sets the read-only attribute, which does not stop new files in a
directory, so the upgrade reached disk and the test failed at its own
fixture check on every Windows CI run since it landed.
Deny write to Everyone with icacls there instead, the mechanism
internal/contract/config's makeDirReadOnly already uses.
Problem: Studio CI caught ArrowDown selection returning to the default row
before Enter could confirm the selected reference.
Root cause: candidates rendered before a passive effect reset selection and
keyboard ownership. That later update could overwrite navigation after the
rows had already committed.
Fix: store the completion answer, active row, and keyboard ownership in one
state value. Publish each answer with its default selection and remove the
deferred reset. Guard navigation against an answer that has become empty.
Verification: four deterministic reference/slash publication and early
navigation regressions fail on the old hook. The PR merge snapshot with this
patch passes 616 frontend files / 5,132 tests; 92 completion/composer cases
pass with --repeats 10. Typecheck, build, census and census verification,
go vet, make lint, make check, and builtin/boot tests pass. Chromium confirms
ArrowDown plus Enter accepts the selected reference and slash command.
Problem: Review requested proof that production completion removes exact
reference candidates independently of the frontend MockPort fixture.
Root cause: The frontend interaction tests and the existing control unit
test did not exercise the registered GET /complete route together with its
controller data, scoped filesystem lookup, and UTF-16 boundary.
Fix: Add seven real HTTP cases using the production handler and disposable
workspace files. Cover partial paths, exact filenames, prose and emoji
prefixes, escaped nested paths, retaining a genuine longer alternative,
and an exact final alternative. No production filtering behavior changes.
Verification: The completion HTTP tests and full serve package pass, along
with go vet, make lint, make check, and the builtin/boot pre-commit suites.
Removing only the existing control.dropTyped call via a temporary Go build
overlay makes six HTTP subtests fail on the retained identity insertion;
this checks regression sensitivity and is not a production defect claim.
Brings the three Web fixes onto the current studio line: the hosted page's own-fetch stale check, the pending approval/ask handed to an attaching polled client, and the no-store shell.
The one conflict, internal/frontend/serve/web_entry_test.go, was both sides appending a test at the same point; both are kept.
Problem: Review requested coverage for Enter after a manually completed
reference, including the case where another candidate is still available.
Root cause: The kernel already removes an insertion equal to the typed
token, but MockPort kept that row and manufactured an extra confirmation
step in development fixtures.
Fix: Keep MockPort's candidate filtering aligned with the kernel and add
four composer regressions. A full path with no remaining candidates sends
in one Enter. An open menu containing a genuine alternative still confirms
that alternative before submission, as required by the interaction contract.
Verification: The three exact-path cases failed against the old fixture.
All 5099 frontend tests, typecheck, build, make check, go vet, and make lint
pass. The existing kernel dropTyped regression passes. Real-browser checks
confirm single-Enter submission for exact paths at wide/light and phone/dark
sizes. The builtin and boot pre-commit suites pass.
A read-only tool that turns a video into a few time-labelled pictures for a
model that reads images. The container is named from the file's own bytes and
ffmpeg is forced to it with file-only protocols; failures carry typed codes.
It is reached through use_capability and stays out of the schema prefix (#13399).
The polled transport attached past the watermark and dropped what came back, so a page opened after a prompt was raised never saw the approval or ask the run was waiting on. An attach now carries the pending prompt to that client alone, and the client reads the frames it is sent.
A hosted page checked for a newer build through the global fetch, which remoteHub has already replaced with a tunnel to the kernel; it compared its own build against the kernel's and reported itself behind forever. The check now reads through the page's own fetch, bound before the swap.
The shell names which build is current and was served with no cache policy, so a reused copy left an open page behind and reloading landed on that same copy. It is now served no-store; the content-hashed bundles under /assets keep their long cache.
With an open-session limit set, opening one more session reuses or closes an
idle pane. A calm note now names the conversation, the limit, and that it is
still in the list; a note is a toast tone with no report button (#13419).
Problem: Pressing Enter while an @ file or directory candidate was visible
could submit an unfinished prompt instead of confirming the reference.
Root cause: The menu highlighted its first item but only owned Enter after
explicit navigation. Directory acceptance also shared the descent path.
Fix: Let the visible selection own Enter and distinguish confirmation from
Tab descent. Finish accepted references with a separator, preserve existing
whitespace and trailing text, and clear the old completion before drafting
continues. Update both locale hints.
Verification: Twelve new composer regressions reproduce the old behavior
and cover files, directories, nested Tab descent, pointer acceptance,
escaped paths, send preferences, Shift keys, IME, and Escape. The full
frontend suite passes (609 files, 5095 tests), alongside typecheck, build,
UI census, make check, go vet, and make lint. Browser interaction was
verified with the local MockPort workspace.
A button on a finished tool card repeats that call's own recorded card in a
section at the end of the transcript, as a view only: nothing is stored or
sent, and the original stays where it happened (#13394).
The transaction-free fallback loaded the disk manifest without the RunID
isolation the locked path applies, so a queue another run (or a previous run)
left behind could be dispatched while the lock was momentarily unavailable.
Refuse a foreign manifest with ErrForeignRun, which the dispatcher already maps
to a retry, and give the frontend a coded refusal for it.
NextQueued returns (InboxItemMeta, bool, error) and, when the disk
transaction is unavailable, reads the published manifest without the lock
(commits land by atomic rename, so that snapshot is complete). The
dispatcher maps a read error to inboxDispatchRetry instead of reporting
the inbox idle.
Narrowed per review: the permanent 1s long stop (armInboxDispatchLongStop,
its constant, its arming and its stop/reset wiring) is removed, so the
retry schedule past the existing bounded burst is exactly what it was
before this branch. The dispatcher audit case is rewritten to assert the
in-scope contract - a failed queue read schedules a retry and does not
report idle, and the durable item survives - rather than delivery after
the burst was exhausted, which was the removed policy.
NextQueued discarded beginDiskTransactionLocked's error and answered from
the in-memory snapshot, so a transaction that could not run — another
process holding the lock, EACCES, ENOSPC, a transient EIO — looked exactly
like an empty queue. dispatchInboxOnce read that as inboxDispatchIdle and
stopped with the durable item still queued, permanently.
NextQueued now reports that failure as an error, and before it will claim
the queue is empty it falls back to reading the published manifest without
the transaction: commits land by atomic rename, so that snapshot is
complete, and repair work stays behind the lock. A genuinely empty or
paused queue is still idle.
The bounded burst (three attempts, ~750ms) used to abandon work its own
scans had seen. Past the burst the dispatcher now parks on a slow
production timer that keeps re-checking until the item advances or the
queue reads empty; reset and Close cancel the pending timer.
A read error maps to inboxDispatchRetry, so an unreadable queue retries
instead of ending dispatch.
When an install already has frontend-next/dist beside the binary, upgrade
reads the page out of the verified archive, validates it in memory, swaps it
in before the binary and puts it back if the binary does not land (#13406).
ListSessionOrder keeps a row whose branch-meta sidecar cannot be read, so one
broken sibling cannot hide a folder, but every column of that row is then a zero
value. recoveryArchiveLineage consumed that row as 'not a recovery copy of this
conversation' and SetSessionLineageArchived reported success while a copy of the
row stayed active and listed on its own. The lineage walk now reads each listing
row's sidecar itself and refuses the transaction, so the archive endpoint
reports the read failure instead of a partial success.
ListBranches dropped a branch whose sidecar could not be decoded while a missing
sidecar built defaults and kept it; interruptedTurnContinuedOnRecoveryBranch
read that absence as 'the turn did not continue on a recovery branch' and
stripped the interrupted turn's messages. An unreadable sidecar is now handled
exactly like a missing one.
Problem: release notes did not mention the focus behaviour shared by all
narrow composer toolbars, making the compact empty state look accidental.
Root cause: the behaviour and pointer stability rationale were documented
only in the internal style contract.
Fix: add matching Chinese and English release-note entries explaining that
empty narrow toolbars remain compact on focus and expand for draft text,
attachments or an open menu, keeping pointer clicks on the same button.
Verification: 12 release-note rendering and Markdown reference tests passed;
make check and git diff --check passed. Documentation-only change.
The cli job builds the page once and packs it beside the executable under
frontend-next/dist, where serve.FindPage looks, through a packer and a checker
that are run by the job and by tests (#13406).
Problem: the five composer menus appeared oversized beside the application's
compact toolbar and other menus.
Root cause: the shared composer menu style imposed 15px labels, 13px notes,
380px panels, 20px corners and generous spacing on every menu. Workspace
toggle and note borders also produced adjacent divider lines.
Fix: use 13px labels and 11px notes, 12px corners and 8px panel insets; choose
280-340px widths by content. Tighten rows, headings, search, chips, checks
and the workspace switch together, leaving one workspace note divider.
Keep all changes scoped to composer menus and update their style contract
and bilingual release note.
Verification: 22 related suites / 226 tests; production build; make lint,
make check and git diff --check. In-app browser checks cover all five menus
in light/dark themes, 390px and 320px viewport bounds, model/branch search,
Escape dismissal and focus restoration. Mock data only.
FindPage now also looks beside the file an executable link resolves to,
after the directory it was invoked from, so a package manager that links the
binary onto the PATH still finds the page next to it (#13406).
Problem: the workspace PR conflicted with the current studio composer, and
restoring its toolbar slot exposed narrow-pane overlap and pointer failures.
Root cause: the new side-question control shares the same toolbar; a fixed
model minimum could overrun its flex group. Focus-dependent phone folding
also moved buttons between pointer press and release.
Fix: retain both worktree setup and the upstream side-question button; allow
model shrinking and internal wrapping at the narrowest widths. Keep empty
phone toolbar geometry stable on focus and extend browser guards to cover
both workspace variants, pointer opening and nested button overlap.
Verification: 5,021 frontend tests; affected suites 126 tests after CSS edits;
production and perf builds; make lint, make check and diff whitespace checks;
hoverradius, branch-menu, modelrows and extended narrow browser guards;
light/dark worktree selection, Escape/focus, menu bounds and first send.
Run analysis and Workbench become icons on the window bar for the focused
pane; the Conversation / Run analysis / Workbench row, the plan peek and
their styles are removed (#13369).
Problem: workspace and branch controls were moved outside the composer, and
the footer-only 8px override broke the shared toolbar radius browser guard.
Root cause: the workspace slot bypassed the existing branch toolbar layout
and overrode the common mode button geometry.
Fix: restore both controls to the original toolbar slot, reuse branch layout
and pill styling, remove the footer overrides and keep runtime statistics
below the card. Cover new and existing conversations in the browser guard,
load the production menu stylesheet in the perf fixture and align the docs.
Verification: reproduced both layout assertions and the light/dark radius
failure before the fix; 4,831 frontend tests passed; production build,
make lint and make check passed; hoverradius, branch-menu, modelrows and
narrow browser guards passed.
A sticky bar of groups (all, new replies, needs your reply, open, closed)
with counts sits at the top of My feedback; rows that do not match are
hidden rather than unmounted so open threads and reply drafts survive (#13388).
The composer's plan fold now carries the /goal text when there is no plan,
and the three pane views get Ctrl/Cmd 1-3 shortcuts and palette entries.
Prerequisite for removing the top plan peek and view strip (#13369).
#13366 bounded the failed tool results a fold keeps, with component tests. This
adds the effect test through the real Build: 60 turns of failing commands fold
seven times, and the request right after a fold must not carry more failed
results than the retention budget plus the recent tail allow (4 with the bound,
9 without it, where the 50% ceiling is the only thing that clears them).
Refs #13148
runSegments read only the final metrics file. When a child was killed or
timed out it folded the missing record in as an accounted zero, deleted the
<metrics>.partial snapshot the sink had been writing, and left Unaccounted
unset, so every published total omitted what the run spent.
Read the sidecar when the final record is absent, keep the file it came from,
set Partial, and mark the run Unaccounted when neither exists.
renderMeterAccounting no longer prices a run whose harness reported no tokens
against the meter, which invented a -100% divergence.
LoadCachedSchemaForSpec rewrote a legacy-keyed entry in memory, discarded the
SaveCachedSchema error, and returned the hit with the new CacheKey. With a
read-only or full cache dir the file still held the legacy key, so boot was
handed a key that a later read cannot reproduce and pinned the session's MCP
tool identity to a key with no on-disk counterpart
(internal/assembly/boot/mcp_registration.go:76).
Keep the hit but claim the upgraded key only once the rewrite reached disk; on
failure the returned object keeps the legacy key the file holds, which matches
the package's best-effort, silent-degradation policy. A successful upgrade is
unchanged, and no caller can now observe a key the file disagrees with.
checkpointImages reported "no such image" for a checkpoint it could not
inspect, so refuseDegradedCheckpointRewrite read a stat failure, a read
failure, a non-regular file, or a file past the replay ceiling as "nothing to
protect" and let a transcript missing an image replace the only copy of that
image — the exact case where the guard cannot know.
Return the inspection error from checkpointImages, and have the guard fail
closed under ErrSessionImagesUnavailable for any checkpoint that exists but
cannot be read whole. A genuinely missing checkpoint still proceeds: there is
no second copy to protect. The restore path stays best-effort and now logs why
it could not inspect the checkpoint.
recordDispatcher.run discarded Writer.Append's error, so Flush — the only
success/failure signal the stats pipeline has — returned nil even when every
queued append failed. The dispatcher now keeps the first failure since the last
flush sentinel and hands it to the waiter through that sentinel, so a caller
that gates a deterministic read on Flush cannot mistake a failed write for a
durable one.
The failure is consumed by the flush that reports it, so a later flush over a
recovered directory returns nil, while a still-broken one keeps reporting. A
full queue may still drop a record; that stays documented and unchanged.
Tests: the audit red test plus controls for a healthy flush persisting records
and for failure-then-convergence.
A fold keeps failed tool results since the latest digest. The projection places
them right after that digest, so the next fold read last fold's survivors as
"since the latest digest" and kept them again: retention grew by every round's
failures and only stopped when the candidate crossed the 50% checkpoint ceiling.
Failed results now have their own retention budget (5% of the window, like kept
user turns), newest first; the rest fold into the digest, whose coverage guard
still names them. [[keep]] messages stay outside the budget.
Refs #13148
The covered-prefix hash includes the system message, so editing the system
prompt between two visits retires a saved compaction. #13336 added the notice
at the agent and control layers; this pins it at boot.Build, and pins that an
unchanged prompt stays quiet.
Refs #13295
Problem: concurrent tasks need separate code directories rooted at a chosen
branch, and composer configuration menus use inconsistent visual scales.
Root cause: isolated workspaces only started from HEAD and their entry was
outside the new-conversation composer. Menu surfaces had independent styles.
Fix: stage worktree and exact base-ref selection before the first submit,
carry draft attachments through confined roots, preserve session settings,
reject stale preparation and roll back unpublished worktrees on failure.
Keep existing conversations and older servers on the branch control.
Share composer menu surfaces, moderate 15px labels, 13px descriptions and
neutral selection states across mode, workspace, permissions, model and effort.
Move workspace controls below the input card beside runtime statistics, retain
branch labels in narrow panes, and wrap model/effort at the smallest widths.
Verification: 4831 frontend tests; TypeScript and production build; real Git
worktree/source/refusal and attachment confinement tests; serve, control,
builtin and boot suites; go vet, make lint and make check. Browser interaction
and light/dark responsive captures are kept outside the repository.
Fixes#13264
A file restore the checkpoint covers only partly stops for one more
answer, taken with Enter or y (the panel's own hint says so); the key
table in both guides listed neither the step nor y. The new test pins
that step: it does not commit before the answer, Esc steps back without
restoring, and Enter or y commits.
agentpreset.Normalize maps "light" and "economy" to balanced, and All()
lists balanced and delivery; light's own policy no longer exists. The CLI
still put light forward as a choice, in seven hand-written strings: the
--preset help of run, tui and serve, the --profile deprecation warning of
each, and parseRuntimeProfile's refusal ("want light, balanced, or
delivery"). The guides went further and described a Light policy (direct-
first planning, targeted verification) that nothing enforces, so a user
who picked it got balanced without being told.
The seven strings now come from agentpreset.All() through livePresets in
preset_flags.go, and the docs (GUIDE, GUIDE.zh-CN, CLI.zh-CN, TOOL_CONTRACT)
offer balanced and delivery, drop the Light policy bullet, and say the
retired names select balanced. light and economy still parse.
TestEffectThinkingTimeIsKeptWithTheTurn failed in the race job on studio
(run 38635511666): "thought kept 46 ms, framed 46 ms; want both at least
50". The scripted provider wrote the reasoning chunk into a buffered
channel and began its 60 ms pause at once, while the host times the
thought from when it takes that chunk; a host that reads the first chunk
late measures less than the pause. Returning the stream 30 ms late
reproduces it (29 ms measured).
The channel is now unbuffered, so the pause starts only after the host has
taken the reasoning chunk and the measured thought cannot be shorter than
it.
acquireLocal waited on the in-process token and ctx.Done in one select.
When both were ready (a lock nobody held, and a deadline already spent)
select picked one at random, so half the time Acquire refused a free lock
with the context error. The file-lock half of acquire already tries once
before it looks at ctx; the in-process half now takes an available token
first too. A held lock still fails at once once the deadline has passed.
This is the failure behind the unsealed evidence bundles on Windows CI.
TestPlanCriterionIsVerifiedByItsStepsCommands (studio run 38664267053)
reported "bundle not sealed: failure trusted_state.unwritable, contract
failure """ after 3.09 s: sealShadowBundle's two-second sealWait covers
the contract seal and object writes as well as the lock wait, so slow
writes spent it, and the bundle's own Append then met a free stream lock
with an expired ctx. With the first write slowed by 2.2 s, 1 of 6 turns
ended unsealed with that exact signature before this change, 0 of 6 after.
Keeping the decoded prefix across ANY decode error dropped every record after
a mid-file break, and that truncated prefix then competed as a checkpoint that
extends the replay. Damage before the last record now refuses the checkpoint.
recallTail had three outcomes and only two reached the audit channel: a
carried memory note and a normal recall each recorded a content-free
decision, while the retrieval ablation switch returned silently and left
lastRecall pointing at an earlier turn. A trajectory could not tell
"retrieval switched off" from "recall found nothing" from "suppressed by a
carried note".
All three now record a RecallResult and emit one memory-recall audit. The
identity travels as a typed memory.SuppressionCode on its own field, from
RecallResult through event.MemoryRecallAudit into the trajectory record's
suppression_code. The prose sentence stays exactly as it was, for readers
only: consumers match the typed field and never split the code out of the
text, so wording cannot become load-bearing again. The new field is additive
and omitempty, so existing records and batches decode unchanged.
The ablated turn also records lastRecall, matching the note path: the
diagnostic surface is defined as the last real turn's decision, and that
turn did decide not to retrieve.
reportBoundaries folded every row into byKey, so a run that never produced a
transcript re-entered the cue-boundary table as a zero measurement after the
funnel, dose and reformulation exits had excluded it. Skip failed rows, count
them, and say scored=0 with the failed count when nothing is left to compare.
A run that failed before its first model turn was appended as a zero-valued
contextMetrics, so results.jsonl gained an empty row, summarize counted it in
Scored and its stage tables, and batch.json reported it as completed with an
empty task id. Stamp such a run with its task, arm and NoEvidence stage, mark it
failed, and keep it out of every statistic while counting it in the manifest as
failed_runs.
load() dropped a turn file that existed but would not read, so a resumed
session numbered the next turn below it and Begin overwrote the only copy
of that snapshot. The store now reserves those turn numbers in memory:
NextTurn() counts them and UnreadableTurns() names them, while Bounds()
and List() still expose only readable checkpoints. No new on-disk state.
A persisted turn-N.json that cannot be read is dropped by load() exactly
like an absent one, so NextTurn() restarts below it and the next Begin
overwrites the only copy. This test operates on a real truncated file
written by the store's own persist path.
checkpointExtendingReplay read the .jsonl checkpoint with the strict loader
and folded a decode error into "this checkpoint does not extend the log".
The in-place append in appendSessionDisplayReadModel is exactly the write a
crash leaves half done, so the torn checkpoint's own intact lines were
discarded with it: resolveReplayedMessages then returned the shortened
replay with err == nil, and the next save rewrote both the checkpoint and
the log from that shorter transcript. Every turn the checkpoint still held
was lost permanently.
Decode tolerantly the way transformAndCopyJsonl already does, and keep the
existing prefix checks, so the tolerant result is adopted only when it
still extends the replayed prefix. A checkpoint that cannot yield a usable
prefix (unreadable, shorter, contradicting) keeps today's "does not extend"
behaviour; there is no third state.
Two gaps in the delivered-fact re-check this branch added:
- applyWrite kept the pre-reload stamps of the facts the model already
holds, so a change that one reload had settled was re-detected by every
later recall: the snapshot reloaded and the whole store was re-tokenized
on each turn until a recall returned hits. Re-stamp the kept facts
against the snapshot just loaded, dropping what it found gone, so the
steady state stays one read of each delivered fact's own file. The
pre-branch code cleared the set on reload, which is why this is not a
new promise.
- The remember tool's typed Memory was discarded at the queue: the note
carries prose and the controller must not parse it back into memory, so
a fact the model wrote itself was absent from the delivered set and an
external edit before the next recall served the stale session snapshot.
memory.Queue now carries the typed facts the tool wrote, remember passes
result.Memory, and the controller records it as delivered. The model's
own write still queues no turn-tail note.
Adds the delivery/change matrix (100 cells over delivery x change time x
change shape) with the reload-count companion, and the provider-boundary
regression for the remember path.
Problem: a fact saved through Controller.SaveMemory and then edited outside the
session is still recalled from the stale snapshot on the next turn; the per-turn
re-check that catches external edits compared each recalled fact with a stat, so
a rewrite that preserved the file's byte length and modification time stayed
invisible and the model kept reading the stale body; and the identity read behind
that stat, Store.Read, resolves a name by listing the whole store directory and
parsing every fact in it.
Root cause: the re-check set held only the previous recall's hits. A save queues
a turn-tail note, and a note suppresses recall for that turn, so a just-saved
fact was never a hit and so was never re-checked. Its check also rested on size
plus modification time, which is not semantic equality -- the repository's
context-projection rule says freshness detection must not rest on mtime.
Fix: baseline the facts the previous turn handed the model -- the previous
recall's hits, plus the facts a write note carried -- with a digest of each
fact's own file, read once at the canonical path its scope and slug name. Before
a recall tail is built those files are re-read directly and their digests
compared: different bytes reload the snapshot, identical bytes never do. The
stamp carries no size and no modification time, so no stat result can decide
that a fact is unchanged; the bytes are the semantic version because they are
exactly what the renderer turns into the model-visible fact. Store.Read stays as
the fallback for a fact whose file is not where its name says.
Cost: one read of each delivered fact's own file per recall, plus one when it is
baselined and, on that baseline, one parse of the same file to confirm the stamp
describes the delivered fact; a snapshot reload only when a delivered fact's
bytes differ or its file is gone. A formatting-only external rewrite re-reads the
fact into the snapshot even though the parsed fact is equal -- the safe
direction. A fact edited into relevance that the previous turn never handed over
stays invisible until the session reloads; no full scan, which
docs/SESSION_MEMORY_RETRIEVAL.md now states, along with the no-hit retention and
the identity fallback.
This extends the merged fix in 99fe3dbbf, which covers the recall-hit path only.
Review round 1 fixed two regressions in this baseline:
- A fact edited between the snapshot load and its first recall was tracked with
the edited file's own digest, so the stamp described bytes the turn never
carried and SameContent() then called the stale snapshot body unchanged
forever. stampFact now marks a fact tracked only when the one file it just
digested re-parses to the very fact being delivered (memory.FactAt + sameFact,
the comparison the untracked fallback already used); otherwise it records the
fact untracked and the identity read decides. The probe stays on that one file,
so no listing and no second scan come back.
- recordDelivered overwrote the delivered set on every recall, so a turn that hit
nothing erased the facts the previous turn was still holding and the next
recall had nothing left to re-check. It now keeps the previous set when this
recall returned no hits, while still appending the facts a write note carried.
Verification: a same-size/preserved-mtime rewrite is asserted first at the
Compose boundary (red on the stat-based check) and then at the true provider
boundary, through the real Build stack with a recording provider that shows the
refreshed body in the request the executor actually sends. The tracked path's
freedom from directory listing is asserted by arming the memory package's
read-dir seam to fail while stamping and while re-checking; FactAt is pinned to
the same single file. Both new tests go red with the stat-based conclusion
restored behind the same API and green with it removed.
TestRecallReloadsAFactEditedBeforeItsFirstRecall (seed, load, rewrite, Compose
twice) and TestRecallKeepsCheckingAfterATurnWithNoHits (recall A, an unrelated
no-hit turn, edit A, ask again) are red on the pre-fix behaviour and green now.
Each review fix was reverted alone with both tests kept: the first test goes red
and the second passes, and the reverse for the other. Go tests, vet, gofmt,
golangci-lint (0 issues) and make check (repolint: clean).
Refs #12948
An explicit expiry is a hard boundary: the fact index and automatic recall drop
the fact past it, and the memory tool paths report it as expired. Pinned guidance
did not check it, so an expired body kept riding the stable prefix, an expired
project fact kept shadowing the live global guidance it used to override, and an
expired body kept consuming the pinned budget that bounds what the prefix sends.
Both places now skip facts that are past their expiry, so the prefix, the shadow
rule and the budget agree with the index and recall. This covers facts that are
already expired when the session loads them; withdrawing a pinned body that
crosses its expiry mid-session is a separate prefix question.
An unqualified save whose name was active in two scopes was refused with a bare
sentence. A code caller could not tell that refusal from a write failure, and
the model received neither the two ids the name could have meant nor a way to
say which fact it wanted.
The refusal now wraps ErrNameActiveInTwoScopes, so errors.Is separates a
cross-scope name collision from a write failure, and its sentence renders that
code, both active ids with their scopes, and the two ways out: update the fact
you meant by its stable id, or save with a scope-qualified name. The remember
tool returns this error verbatim, and the agent renders it as the model-facing
tool result, so the identity and both ids reach the model without a second,
drifting sentence.
Tests cover the identity and both ids at the store and at the tool, the refusal
leaving both facts at their own bytes and revision, and the two unambiguous
saves (stable id, scope-qualified name) still succeeding. A neighbouring test
pins that a name active in exactly one scope is still updated, not refused.
An unqualified name that was active in another scope was archived after the save
was published, so a failure there left the same name active twice, and the
archiving itself removed a fact the caller never named.
The collision is now reported before anything is written, and the post-commit
archiving is gone.
git diff --numstat -z writes a rename or copy as the counts, a trailing
tab, then the old and new paths as their own NUL fields. Trimming that
record removed the tab, so the parser treated it as a short line and
dropped it. Status and Summary then reported no line counts for the
change. The counts stay on the new path, which is the name status lists.
git diff --numstat -z emits a rename or copy as the counts, a tab, NUL,
the old path, NUL, and the new path. Status and Summary need those counts
on the destination, beside an ordinary edit in the same diff.
The CLI/TUI shortcut tables in docs/GUIDE.md and docs/GUIDE.zh-CN.md still
described four 1.x behaviours the 2.x terminal UI does not have (#12144
lists them as missing):
- Up/Down "navigate queued follow-up feedback" while a turn runs; in 2.x
they recall history there too (recallAtEdge has no running case), and
queued items are managed with /queue.
- Esc "un-sends a just-submitted turn before any reply"; escape() cancels a
running turn, clears input, or leaves ! shell mode, and nothing un-sends.
- Composer text selection by drag with replace-on-type; onMouse ignores a
left click below the transcript viewport, so only transcript text is
selectable, and Ctrl+C copies only that.
- The rewind picker's f (fork); rewindActions has b c d s u.
Two tests pin the rows as now written: Up recalls while a turn runs, and
Esc on an empty composer leaves shell mode.
TestAuthorizeHTTPMCPDoesNotHoldStateLockDuringBrowser failed on Windows
CI (studio run 38416026313) with "removal during browser flow blocked on
OAuth state lock" after 1.09 s. The test runs ReconcileHTTPMCPOAuthAfterRemoval
from inside openURL and gave it one second. AuthorizeHTTPMCP holds no lock
there (captureMCPOAuthGeneration releases its lock before the browser
opens), so a slow removal on a loaded runner read as a held lock. Starting
the removal 1.2 s late reproduces the message.
The wait is now testenv.Budget(t). A lock really held across the browser
is still caught: the removal's own lock wait gives up after five seconds
and returns an error, which fails the test.
TestSaveSnapshotSucceedsWhenCrossProcessFileLockReleasesBeforeDeadline
failed on Windows CI (studio run 38502171680) with "lock session file:
session file lock held" after 1.28 s. A goroutine releases the competing
lock after 30 ms and the save waits at most 500 ms for it; on a loaded
runner the release came later than that. Releasing after 600 ms
reproduces the message.
The save returns as soon as the lock is free, so the wait now is
testenv.Budget(t). TestSaveSnapshotBoundsCrossProcessFileLockWait keeps
its 40 ms, since it asserts that a lock never released makes the save
fail.
TestHostContextCarriesTheExtraWritableFoldersToLocalServersOnly and
TestHostContextCapsHowManyWritableFoldersAreTold fail on every Windows run
since e6f70622b (studio run 38537039922): their roots are written as
"/data/shared" and "/data/000", which have no drive and so are not
absolute on Windows, and metaWritableRoots drops every root that is not.
The key is never sent and both tests see nil.
The roots are now built from the temp directory's volume, the repo's
usual way to spell an absolute path on any OS. The same change makes
TestHostContextSaysNothingOfWritableFoldersWhenThereAreNone exercise the
workspace exclusion on Windows: its "/w" was dropped as relative there,
so it passed whether or not the workspace was excluded.
Add a separately selectable naming model and adjustable reasoning preferences
for chat roles, with typed roles/efforts responses and user-only role settings.
Keep title generation fixed off with temperature zero and a 60-token budget;
reject explicit title effort instead of storing an unused setting. Use the
configured startup chat model when no naming model is selected, including
servers without a running controller.
Integrate the on-demand naming and rename paths from #11591 with grouped title
provider state. Preserve default-model persistence, per-profile model overrides,
busy-save notices, and inherited effort validation against the executing model.
Share the model-preference row's existing layout declarations to keep the CSS
line count below the studio baseline without changing rendered layout.
Preserve each provider's saved effort when a role has no stored preference;
use the model default only for explicitly selected auto. Add a real boot
assembly effect test for planner and guardian-derived reviewer/evaluator
providers, while leaving subagent effort inheritance unchanged.
The title-bar Browser button and the Workbench tab opened the same column but
the button also asked for a blank browser page. The tab is now the only entry.
The buttons sit left of the rail, and the shut dock's tab sits flush with the
window edge instead of over the rail's ticks. The browser guard asserts no two
of the buttons, rail, tab and composer intersect at four widths.
Windows answers "not found" for a file where a directory should be, or
beneath one, so the not-exist shortcut hid the very case the directory read
error exists for. The stat of the entry and its parent tells them apart.
Problem: Studio user-message and composer images only showed thumbnails,
while tool screenshots used a separate enlargement overlay.
Root cause: The three entrances did not share a preview interaction owner.
Fix: Use a common image trigger and body-portaled viewer with original
sources, keyboard access, focus restoration, safe double clicking, explicit
closing and unavailable-image handling. Remove the tool-only overlay and
document the restored attachment parity with main-v2.
Verification: 541 frontend test files and 4611 tests pass; typecheck,
production build, census verification, make lint, make check and go vet
pass. Builtin and boot Go tests pass. Chromium checks cover light/dark,
wide/narrow windows and 150% interface zoom. The browser guard is wired
into Studio CI. Correct the perf fixture to await committed language state.
The wire log records whether a surface frame was dropped; the surfaces route
answers 200 with an incomplete mark only then, and an unreadable history has a
code of its own.
Problem: permission and storage failures can hide existing memory revisions,
make restores report them as missing, and compute recovery revisions from an
incomplete history.
Root cause: directory discovery conflated every ReadDir error with absence.
Fix: unify all eleven store walks behind a typed reader that treats only
missing directories as empty. Propagate errors through reads, save guards,
recovery, recall, controller commands, and HTTP responses. Complete recovery
and legacy duplicate discovery before writing. Preserve persisted formats.
Verification: failing-first permission fixture; eleven-site permission,
I/O, and non-directory regressions; unchanged files on failed recovery or
save; missing-directory compatibility; tool/controller/HTTP and bilingual
error coverage; Go tests, vet, lint, repository checks, frontend tests and
typecheck, and Windows memory test cross-compilation.
Fixes#12950
Problem: the pending icon restoration was missing from the next-release
notes, so users would not be told why the application artwork changed.
Root cause: the original restoration commit only included icon assets.
Fix: add a bilingual release-note fragment covering the application,
taskbar/Dock, launcher and tray icons without changing any artwork.
Verification: the 16 release-note rendering/publication tests, repository
lint, Go vet, changed-file checks and diff whitespace checks pass.
Problem: Studio 2.39.0 replaced the expected black-and-gold application,
taskbar, Dock, launcher, and tray artwork with blue-and-yellow icons.
Root cause: the shared SVG and all generated platform resources were
recoloured together in #12897.
Fix: restore the black-and-gold SVG and its ten generated icon files,
preserving the platform crop, ICO/ICNS sizes, and Windows tray DPI scales.
Verification: regenerate all assets and compare their bytes with the
pre-#12897 artwork; decode and visually inspect all 26 raster sizes;
run the Electron unit suite, Go vet, repository lint, and changed-file
checks, plus the builtin and boot suites. No application or packaging
logic changes.
The guard only saw byte-periodic blocks. A second structural reading of the same
tail - many short lines from very few distinct ones - catches a degraded stream
that never repeats a block.
The program find runs decides: a bare read-only name whose own arguments do not
write. Anything else, -ok, -delete and the -fprint family stay writers, so a
read-only find no longer takes the workspace write lease.
A bare /theme opens the quick picker; moving repaints, Enter keeps and stores the
row, Escape restores the original. /theme list keeps the old listing.
A shadowFloat token with a strict px/hex shadow grammar maps onto
--shadow-float, and onto --shadow-pack for the light menus that kept their
own default. Unset, nothing paints differently.
Two tests for the path #12966 settled: removing every provider saves an
empty list that loads back empty, and adding one afterwards replaces it.
Over HTTP, removing the last provider moves the runtime to the stand-in
and lists no provider.
Both fail without the providers = [] render line from #12966.
Refs #12915
Declares import_web_image's snip stance (a small JSON receipt takes the default) and regenerates
docs/TOOL_CONTRACT.md, the two contract tests that landed on studio after the branch point.
move_file now asks for read access to its source on both paths (a forbidden-read source is refused
before the rename, not only when the copy fallback runs), and four agent tests built after the
studio branch point supply the permissive target-access callback the agent now requires.
Keeps the row-level double-click rename from studio (and drops the title-level handler it replaced),
declares the dialog's title input as session.rename.title, and expects the inline mode in the
double-click test now that onRename carries it.
Keeps the new summary text, names the advanced-appearance disclosure in the settings catalogue (the catalogue test requires a titled block to name itself) and folds three selector lists onto one line each so app.css stays inside its recorded size.
Folds the studio's retired-surface rows into the identity merge of a rebuild (rebuild.ts), and lets two
fixtures report the turn as running while it runs, since a status read now vouches for it.
The same HttpError check and the same note block were copied into Compaction, Rules, Shell, Context,
Settings and EditConn. savedNotApplied lives next to SAVED_NOT_APPLIED and SaveNote draws the note once.
Mantine's gray primary filled colour is #868e96, which is 3.4:1 against the button's white label.
The scoped provider now maps it to the app's text token.
In a narrow settings column the extra control pushed the scope and switch of the longest rows past the card edge. Next to the name it keeps the other columns where they were.
Keeps one /answer handler that serves both question engagement and surface answers, and
carries the computer driver and native-surface options through the workspace switch.
repolint's process-input rule flags a kernel package that reads SystemRoot. GetSystemDirectory
is also the stronger answer: a changed environment cannot stand in for System32.
List answered "no distributions" for any wsl.exe failure, so a probe that hit its
deadline or was cancelled read as a machine without WSL. The caller's ended context
is now returned as such.
Persist undo retirement before admitting new turns or tool writes, and refuse new activity when that persistence fails. Keep post-tool checkpoint-recording warnings separate from truthful tool results. Preserve the preceding undo after fully compensated rewind failures, and block activity until incomplete recovery settles. Reuse the existing Studio notice, availability route and change event; retain the failure-path tests.
A conversation whose log still lives in the 1.x store is read here and never
written, so it shares the sidebar with this build's own rows and cannot be told
apart. The tree now reports which store each row came from (the kernel decides
it through IsForeignSessionLog) and the row shows a small "Old" mark at its end,
which yields to the hover actions.
The phone-access panel renders data-action="share.remember" and reads
hub.shareRemember(), but neither reached the action registry nor the test
hubs that build one by hand: actions.test.ts failed with "rendered but not
in src/actions.ts" (PhoneAccess.tsx:194), and PhonePop.interaction /
PhoneAccess.port failed with "hub.shareRemember is not a function".
Register the action beside the other share.* entries and give both test
hubs the hook the panel now reads.
A phone pairs by scanning a QR code, and the credential it keeps rides a
cookie scoped to the origin it scanned. The host held that pairing in memory
only: closing the share or restarting the process threw every device away, so
the next launch asked for a fresh scan.
[serve] remember_paired_devices, false by default and the user's alone (a
project reasonix.toml that sets it is ignored and reported), keeps three things
when it is on: the phones that paired (digests only, written to
<state>/device-trust.json and adopted at startup), the address the share last
listened on, and whether the share was open at all. The persist hooks read the
key when they run, so the panel switch takes effect without a restart, and
turning it off forgets on the spot: everything paired is revoked and both files
are removed. A snapshot taken before a later change is dropped by revision, so
a concurrent pairing write can never overwrite a revoke. Address ordering keeps
its upstream preference. A phone that paired never expires: the point of the
switch is that one's own LAN phone connects again next time, and that is
intentional.
Closing the share still unpairs everything; that is the person withdrawing it.
Only the process going away keeps the phones. The state directory is the
kernel's, outside every workspace, and agent-run commands have no path to it:
sandbox writes are confined to the workspace and ~/.reasonix, and the trust
file never enters them. A restart reopens the listener only when the switch
was on and the share was open when the process left — the state of things the
person chose, not a new one.
Fixes#11834
Refs #11878
Problem: frontend CI intermittently asserted that Files was hidden before a
fresh browser-tab follower had committed its selection.
Root cause: retiring the preview's request record does not settle the separate
Promise.allSettled callback that updates the visible workbench selection.
Fix: wait for the asserted workbench DOM state, as the sibling preview test
already does, while preserving the selected-tab and newer-intent guarantees.
Verification: all 49 HTML preview and neighbouring browser race/rejection tests,
frontend typecheck, and repository lint/check pass; prescribed Go preflight
checks run before commit. CI is monitored on the pushed head.
TestE2ESessionLoad failed on Windows CI on studio (run 38179453323) with
"session/prompt: timed out" after 7.64 s: drainPrompt gave up when five
seconds passed without a notification or the prompt's response. A full
prompt crosses the ACP server, controller, agent and transcript
persistence, and a loaded Windows runner can stay quiet that long without
anything being stuck. Pausing the scripted tool for 5.5 s reproduces it.
drainPrompt now waits rpcCallBudget(t), the testenv.Budget wait the same
file's other RPC helpers already use.
Problem: The latest activity-header note conflicts with this PR's sidebar
entry at the end of the unreleased notes.
Root cause: Both branches append their independent entries at one anchor.
Fix: Preserve all upstream notes and place the sidebar entries alongside the
session-list entries. Retain the narrow-only drawer control and the sticky
activity header in the auto-merged stylesheet.
Verification: Sidebar/divider, responsive layout and activity-fold tests;
the Chromium sticky-header guard in light and dark; production build,
repolint, release-note/reference tests, typecheck and UI census.
TestOnlyChangedPathsAreAcceptedAndTheyAreLiteral fails on every Windows
run since 40e097879 (studio run 38383059323): it writes a file named
"*.txt", and a Windows file name cannot hold "*" ("The filename,
directory name, or volume label syntax is incorrect.").
The name now is "[ot][nw][eo].txt". As a git pathspec glob it matches
one.txt and two.txt, as "*.txt" did, and every filesystem accepts it, so
the test still proves Stage passes each path as :(literal) on all three
platforms.
Problem: The latest studio session-list release notes conflict with the
sidebar-control entry in this PR.
Root cause: Both changes prepend their notes at the same unreleased anchor.
Fix: Preserve the upstream session-list and theme notes, append the sidebar
entries in the existing flat-list format, and retain narrow-only drawer
controls alongside the upstream spacing changes.
Verification: Sidebar/divider, session-list, remote-title and spacing tests;
production build, typecheck, repolint, UI census and fixture verification,
and release-note/reference tests.
A turn that already carries a memory note skips automatic recall, because the
note says the same thing better. That decision was recorded for the session's own
diagnostics but never reached the audit channel, so a trajectory could not say
that a real recall was suppressed and why.
The suppression now goes through the same content-free audit as a recall that
returned hits. A turn whose retrieval is disabled by ablation is unchanged.
The plain save path, which covers a new fact and an update that keeps its name,
wrote the fact and then published the index. A failed publication returned the
error with the fact already written, so a retry of the same intent either
advanced the revision again or was refused by the revision guard the caller had
just satisfied.
The write is now withdrawn when the publication fails: a new fact is removed, an
update is rewritten from the bytes and mode that were on disk — not re-rendered
from the parsed fact, which would silently drop frontmatter this code does not
know, comments and hand-written formatting. If the withdrawal itself fails the
error carries the state it left.
Problem: the brand patch conflicted with Studio and failed the sidebar and
type-scale contracts; the empty-session symbol replayed an entrance on mount.
Root cause: the presentation change retained old fixture expectations, typed
subtitle sizes directly, and treated a mounted brand as a new event.
Fix: integrate the current Studio base, retain its sidebar controls, update the
wordmark assertion, reuse shared size tokens, keep the hero symbol still, and
document the final brand contract without stylesheet commentary.
Verification: frontend typecheck/build/census, affected tests, Chromium guards,
go vet, make lint/check, and builtin/boot tests passed. Full frontend validation
is recorded in the PR; CI is evaluated on this merge's head before delivery.
Problem: Studio packages need payload and installed-shell acceptance, and a
complete artifact gate before publication. Running the Electron and matrix
contracts on Windows also exposed host-dependent test fixtures.
Root cause: #12914 supplies native Linux ARM64 packaging but checks only Go
payload architectures and selected artifact names. The macOS bundle fixture
used the host OS's temporary path, and the matrix fixture relied on PATH
lookup for a mock git command instead of binding it in the tested shell.
Fix: Keep upstream native runners and system FPM, consolidate ELF checks,
verify both Linux package formats, exercise the installed shell, and require
all twelve nonempty release artifacts. Guard installation acceptance with
GITHUB_ACTIONS=true. Use platform-specific workspace fixtures and a shell
function for the matrix's git fixture, keeping all Windows test coverage.
Rebase on current studio and preserve adjacent release notes.
Verification: Electron 143 tests and Linux/script 39 tests pass, along with
workflow contracts, desktop Go tests, builtin/boot regressions, vet, lint,
repository checks, shell syntax and Actionlint. Previous-head hosted native
Linux package verification and installed acceptance passed for amd64/arm64;
replacement-head Windows CI remains the platform acceptance gate.
Refs #12906
Problem: The sidebar-control change conflicts with the theme release notes
added to studio after the PR was opened.
Root cause: Both branches append their next-release entries at the same
location in the unreleased notes.
Fix: Preserve both sidebar and theme entries, use the accepted release-note
heading, and keep the bilingual sidebar entries within the prose budget.
The auto-merged stylesheet retains narrow-only drawer controls and current
theme foundations.
Verification: Full frontend tests, typecheck, production build, repolint,
UI census and fixture verification, and release-note/reference tests.
Problem: PR #12935 became conflicting after the Studio palette changes landed.
Root cause: Both branches appended bilingual entries to the same unreleased
notes, and both touched the shared Studio stylesheet.
Fix: Merge studio while preserving the session-list and palette release notes.
Keep the upstream theme tokens alongside the compact session-row rules.
Verification: gofmt, go vet, make lint and make check passed. Frontend typecheck,
production build, UI census, all 92 focused session and theme tests, and the
light/dark 34px row geometry guard passed. Builtin, boot and theme Go tests
passed. Hosted CI still requires upstream workflow approval.
Problem: The session sidebar mixes one- and two-line titles and reserves too
much vertical space. The persistent local-selection entry adds another blank
row between the filters and the machine tree.
Root cause: Studio clamps titles to two lines with a 42px minimum row height
and vertical margins; SessionBatch renders its entry even outside selection.
Fix: Use 34px single-line rows with full-title hover hints, move local batch
selection into the machine menu, and restore focus to that menu on exit.
Keep batch archive/delete and neighbouring row actions intact. Add browser
geometry coverage to Studio CI, interaction regressions, parity/release notes,
and before/after screenshots for light/dark wide and narrow windows.
Verification: Go formatting, vet, lint, builtin/boot tests, frontend typecheck
and build, UI census, browser geometry and affected interactions passed.
The frontend full run passed 4156 of 4157 tests; the unchanged HTML-preview
failure passed all 26 tests on isolated replay. The optional translation scan
reports the same two missing test-only keys on the base and this branch.
Three control tests failed on Windows CI at a fixed one-second wait for
something the controller does asynchronously:
- TestInboxDispatchKickDuringEmptyScanIsNotLost (run 38178550657) and
TestRepeatedJobCompletionIsIdempotent (run 38283959118): "timed out
waiting for inbox dispatch" from waitForInboxDispatch.
- TestEditPromptPersistsOriginalPrompt (run 38367692250): "edited prompt
was not persisted before deadline" after its one-second poll.
Delaying the scripted runner or the provider by 1.2 s reproduces all
three messages. These waits only catch a hang, which is what
testenv.Budget is for; every one-second wait in inbox_dispatch_test.go
and the edit-persist poll now uses it. The 150 ms waits that assert
nothing happened are unchanged.
Problem: Studio offers no plain built-in alternatives matching the white/blue
and warm/tea-green directions requested in #12909.
Root cause: the default swatch inherits active pack colours, plain packs lose
the default reading surface and weight, and legacy Studio rules override
semantic accent colours on selected controls and dark send buttons.
Fix: ship two light/dark token-only packs, share canonical default preview
colours, preserve plain reading styles, and close the affected controls over
theme tokens. Add optional accentInk, localized labels and backdrop grouping.
Verification: 209 frontend tests, production build/typecheck, theme and HTTP
server tests, go vet ./..., make lint, make check, builtin and boot tests.
Browser checks cover system-following, default restoration, and light/dark
workbenches at desktop and phone widths; screenshots are included.
TestSurveyReturnsPartialCountWhenAWalkStalls failed the coverage job on
studio (run 38341178074) with "the partial count must survive: 0 bytes,
0 files". It gave the walk a 50 ms budget and expected it to count the
one file before running out; the walk's callback checks the budget's
context first, so a walk goroutine scheduled late counts nothing. A
100 ms delay before the walk reproduces it, and the same delay fails
TestTruncatedRootKeepsTheErrorFoundBeforeTheBudgetRanOut, which relies on
the same 50 ms.
Both now use stalledWalkBudget, 500 ms: ten times what the failure needed,
while Survey's three stalled roots still return well inside the test's
5 s guard.
TestReaderIntentRefusesDispatchAfterSafetyDrift failed the race job on
studio (run 38334190859): the test replaces rt.schema at
plugin_test.go:1765 while the handshake's detached cache write
(saveHandshakeSchema -> cacheableToolsOf, cache.go:44) can still be
reading it. Nothing in the product writes a remote tool's schema after the
handshake; only this test does.
The test now drains host.bgWrites, which every detached writer joins,
before replacing the schema. With the cache write delayed by 300 ms, the
cached schema was the test's replacement before this change and the
handshake's own after it.
The commit button only records what is staged, and Studio had no way to stage:
an empty index sent people to git add. The changes list now shows each file as
staged, partly staged or not, ticks and unticks them, and stages everything at
once; the kernel accepts only paths git itself lists.
Problem: The desktop sidebar has a redundant top hamburger button alongside
its divider, making the sidebar affordances ambiguous.
Root cause: The drawer entry is mounted in all layouts, although desktop
layouts already expose a divider that toggles and resizes the workspace rail.
Fix: Mount the top entry only under the scene fold, remove its desktop chrome
spacing, and preserve the existing divider and remembered width. Document
the intentional control removal and include light/dark responsive screenshots.
Verification: 488 frontend test files and 4141 tests pass, including divider
interaction and responsive boundary coverage. Typecheck, production build,
repolint, UI census and fixture verification pass. Browser interactions and
an isolated macOS Electron window verify the retained controls.
Refs #12917
Add linked maps, comparisons, dashboards and workspace-backed photo acquisition
and viewing for desktop task surfaces.
Certify Mantine viewer interactions and isolate Windows command-line and direct
DNS test fixtures from host process/proxy behaviour.
Cache impact: shortened write_file schema changes tool/prefix hashes once on
upgrade (5572 to 5524 schema tokens); system prompt stays unchanged. Update the
boot goldens for that deliberate contract change. Image discovery stays hidden.
The release builds linux/arm64 on a native ubuntu-22.04-arm runner and requires
all four Linux packages by name. electron-builder's own fpm is x86_64-only, so
the arm64 leg installs one from RubyGems. The package job checks that the Go
binaries inside the .deb are built for the architecture the package is named for.
The status channel of a child's progress rides ToolProgress, which the
broadcaster treated as droppable. Its terminal phase has no successor, so a
backlogged client lost it and kept showing the child as running.
A click on a Windows toast grants the right to take the foreground to the
process it activates, which is the kernel host; the window to raise is the
shell's. The host now grants the right before forwarding the click, and the
shell lifts a window the platform refused to focus. A toast that could not be
delivered is written to the host log instead of being dropped.
Machine rows become small section headers with counts in words, project rows keep
the strongest line with a numeric pill, scope filters hide zero counts and read at
12px, and the revealed + and ... buttons are outlined and share one hover, press,
expanded and busy state.
The icon keeps the R-and-whale silhouette and drops the gold gradients for flat
colour: blue tile, yellow R, a blue whale with a pale belly, eye and cheek, and a
band of waves. The sidebar brand mark is the same file instead of a drawn glyph.
A reply to a fixed report means the fix did not hold. feedback-sync reopens an
issue closed as completed after mirroring the reply, and Studio offers the
reply box on fixed reports.
Switching views ran a view transition over the whole pane, so the tab bar was
cross-faded with it: the old tab's box faded out while the new one faded in,
read as two separate animations. The highlight is now one element that sits
under the chosen tab; during a switch the tab bar is not cross-faded and the
highlight slides and resizes to the new tab, with the labels above it.
MigrateV2 rewrote a legacy fact as V2 and then published its index line. If that
publication failed the migration reported the error with the fact already
rewritten, and the next run saw a fact that already had an id and a revision and
skipped it, so the index stayed stale with no path back.
A fact that already looks migrated is now reindexed instead of skipped, so the
second run converges on the index the first run could not publish.
The kernel answers POST /browser/close with 204 and no body, and the port read
it as JSON, so every successful close reported "Unexpected end of JSON input".
wait now echoes the jobs' new output, as whole lines and capped at 64 KiB,
through the call's live progress, starting from their latest lines. The
frontend folded a progress frame like a dispatch, so the frame's empty name
blanked the step and each chunk replaced the last; it now appends the chunk
and keeps the call's name.
--dock-w was a registered, inherited custom property tweened on .app, so every
frame of opening or closing the workbench restyled every node in the window.
It now does not inherit and lives on the pane whose grid reads it; the gutter
writes it there while dragging. Style recalculation while switching halves,
and the long frames on a throttled long transcript mostly go away.
With the display currency left automatic, the server adopted the wallet's
currency for its own ledger only: the quotes on usage events and the frontends
were never told, so the status bar showed this turn in US$ beside a CNY wallet.
The controller's Balance now adopts it and announces it as an operator notice,
the same path a chosen currency takes, so every quote and frontend follows;
a chosen currency still wins and nothing is written to the config.
The folder picker also shows hidden folders, where tools keep their state.
Enforce workspace read policy before previews, diagnostics, hooks, and file execution.
Provide admitted previews to permission approval and Auto Guard, then capture fresh preimages after PreToolUse. Share target access checks across execution, planning, review, and delegated agents.
Keep genuinely write-only targets available, report existing-file overwrite refusals with write-specific messages, and avoid hook-write coverage gaps for read-only calls.
Preserve network-path protection before filesystem lookup while allowing authorized network roots and registered folder references.
Include baseline-compatible disclosure regressions, production target-access wiring checks, and localized tool-card tests.
Windows 11 draws a frameless window's border from the frame's scheme, which
Electron takes from nativeTheme. Studio never set it, so a dark theme chosen in
the app on a light system kept a light frame, seen as a light line around the
window. The frame now takes the stored theme (auto still follows the system),
at launch and whenever the preference is saved.
A project row put its session count on a second line, so every project took
two rows. The count now sits where the row's actions are and gives way to them
on hover or keyboard focus, so a row is one line and the name keeps its width;
touch screens show the actions. A tag still takes a second line. The two add
buttons share one size and hover, and the filter labels and counts are larger.
The docked workbench card sat 8px into the pane with a 39px strip, so its
bottom line ran 7px below the conversation's tab row and the two read as
unrelated bars. The strip is now the shared tab-row height less the card's
inset and border, so both lines meet; the card keeps its corners and edge.
Settings, feedback, a full screenshot, a checkpoint preview and a sub-agent
transcript open in veils that start below the title bar, so centring inside
them put each dialog half a bar below the window's middle. Each veil now
carries a transparent bottom border as tall as the bar, which makes the box it
centres in symmetric; the phone layout keeps settings full screen.
[agent] host_verification = "off" in the user file drops the check, review and
sign-off a turn that changed files owes before it may finish. The turn's own
unfinished plan and a capability the user required still hold it, and the
receipt still says what was and was not checked. A project file can neither
turn it off nor undo the user's choice.
Escape stopped a running turn at once, so a stray press ended a long task. The
first Escape now arms the stop and shows a hint for 2.5 seconds; a second one
inside that window stops. Appearance gets a per-machine choice to stop on the
first press again. The stop button is unchanged.
openWorkspace now opens a folder only when it is a directory and, on macOS,
not an application bundle; anything else is selected in its parent as before.
With the icon rail off or collapsed the whole sidebar head was indented 44px
further so the toggle had room beside the wordmark, which left a blank strip
beside the new-session, search and quick-entry rows. Only the wordmark row
makes room now; every other row sits at the same inset in all three modes.
The toggle was declared before the nav column, the sidebar head and the
chrome, all of which are window-drag regions in the desktop shell. Where drag
regions overlap, the one declared later wins, so a click on the toggle moved
the window instead. It is now the last child of the app; its z-index drops one
step so the settings sheet still covers it.
A selection reaching into blocks that are not mounted mounted every block in
between in one synchronous render, so a drag across a long conversation froze
the window (22s at 1200 turns). A block the selection only passes through, with
neither end inside it, is now held as the conversation's own text at its
placeholder height; blocks with an end inside stay whole.
A tool whose prefixed name exceeded 128 characters made the provider refuse
every request in the session. toolName now shortens only the tool part,
keeping the server prefix and a hash of the full name so siblings stay
distinct and the name is deterministic for permission and hook rules.
The refresh listed and built its own catalog, so it reached the model with
definitions the pin judgement #12597 added had never seen: a server could wait
until the session was running and announce a tool nobody approved. The judged
path now has one entry point for both listings — read, then publish — and the
refresh goes through it.
readToolCatalog takes no lock and publishes nothing, so the refresh keeps
building its candidate off the lock and swapping it in as one piece, and
publishToolCatalogLocked moves every field together: what was withheld was the
one a caller could forget, and it is the one that makes the judgement visible.
The digest covers the held set, so approving definitions the server had added
is itself a change the next listing publishes.
What the session does with that follows: the registry's held set is replaced
from the same judgement, as at every other registration site, and a pinned slot
whose tool is held says so instead of claiming the server dropped it.
hostNotices moves to sessionRuntime, which is its lifetime: a withdrawn slot
outlives the conversation that saw it. That also keeps agent.go inside its
file-size budget, and the catalog path moving to tool_catalog.go keeps
plugin.go inside its own, with no baseline spent.
Refs #11957
A server that drops a tool it had pinned into the provider-visible array was
rewriting that array mid-session: the cache-stable prefix moved and every token
behind it was spent again. The entry now stays until the session ends, the call
it still invites fails with the live catalog's not-found, and the model is told
on the turn tail — the one place an append leaves the prefix byte-stable.
Registry.ReplacePrefix swaps a server's namespace under one lock. The removal
and the re-registration were two calls, so a turn composing its request between
them was built from a server that appeared to have no tools.
A server that declared tools.listChanged may add, rename or drop tools after
startup and says so with notifications/tools/list_changed. All three transports
dispatched notifications/progress and dropped every other notification, so the
catalog stayed as first listed: a tool the server added was unreachable, and one
it removed was still offered to the model.
The transports now surface a notification by method, the client re-lists once a
burst settles, and the Host publishes a catalog that actually changed. The
session's registry re-registers it, so what the model can reach follows without
moving the cache-stable prefix.
Registrations live on the reconnecting transport and are rebound onto a
replacement connection, and a backend rolled in behind the stable proxy is
watched like any other, so a connection that dies does not leave the server
unheard for the rest of the session.
The unrecognised-write effect test asserts on what the turn's commands
changed, but left the sandbox at its default. A runner that has no bwrap
refuses to run them at all, so the case that keeps the waiver owed the check
for a reason the test is not about: green on the `test` jobs, which install
bubblewrap, and red on `coverage`, which does not.
Every other effect test that needs a shell writes `[sandbox] bash = "off"`;
this one now does too. Reproduced with a PATH that hides bwrap, which fails
with the CI's exact line and passes with the fix.
A turn that wrote code and then ran any command the classifier could not read
as a check was released as verified. The waiver was keyed on one fact — "not a
check" — while the receipt carried a second one it never read: whether the host
watched the command write. `echo … > generated.txt` after a code write is
graded MutationProven from the redirect itself, and the turn still finished
with the change unverified.
The waiver now also requires the grade the host already recorded, and requires
it of every unrecognised command after the write rather than the most recent
one: a quiet `ls` beside a writing command says nothing about that command. A
turn whose unreadable commands all left the tree as the host found it keeps the
waiver — that miss is the classifier's blind spot, not the turn's. A write the
host watched, and a change whose extent it could not establish, leave the check
owed: an incomplete observation never establishes scope, and unproven_mutation
is settled by observation rather than by a check that ran afterwards.
Ordinary turns only: Delivery gates verification through a corroborated
citation and never had this waiver.
Refs #12567
Two tests that read a turn's sealed bundle have failed on Windows CI with
nothing that says why the bundle was missing:
- TestContractCriterionIsSatisfiedByItsOwnCheck (runs 38185765331 and
38199681271): "trusted state tampered: malformed digest """, because the
test read the empty Record of an unsealed audit as a digest.
- TestEffectSupportingWritesWithoutChecksMayFinish/markdown_only (run
38252045191): "missing sealed evidence".
The audit already carries the store's typed cause. sealedRecord fails with
FailureCode and ContractFailure when the bundle was not sealed, and every
agent test that reads a sealed record goes through it; the boot test prints
the audits it found.
TestSessionViewedToleratesASidecarCaughtMidReplace failed in the coverage
job (run 38264909143) with "a view that met a torn sidecar failed instead
of waiting it out" in 0.20 s. The view retries a torn sidecar for 20, 50
and 100 ms; the test heals it from a goroutine after 30 ms, so the heal had
140 ms of slack, and on a loaded runner the goroutine started later than
that. Delaying the heal to 250 ms reproduces the failure locally.
The test now appends a 5 s last retry to the product's schedule, the way
the meta-ledger tests already replace it, so a late heal is still waited
out. A reader that does not retry at all still fails the test.
Pinning a conversation only fed the Pinned filter and the menu label, so in the
full list nothing moved and nothing showed it was pinned. Pinned conversations
now come first within their project, the kernel's order holding inside each
group, and each carries a small pin.
CopyButton gained both a menuitem role and an aria label on studio, the
actions list gained reply.branch, and the new selection tests build a reply
action set that now also needs the branch pair.
A replacement published beside the old tree keeps the old tree, so an
uninstall that removed only the active root left the managed directory
behind. Removal now also deletes the canonical copy and earlier staged
copies, skipping any directory the registry still points at.
rewind_changed is a content-free invalidation a reopened pane replaces with
a read of /rewind/undo, so the wire log skips it, and the new close button
declares its action id.
A server whose tool definitions differ from what the user approved has those
tools withheld from the model, and the kernel already had an approval route
for them, but the settings row never said so and offered no way to approve.
The row now carries the withheld tools with how each differs, and a button
that sends back the digest of exactly that set, then reconnects.
The page's iframe now exists only after a navigation and the browser opens
from the workbench launcher, so the tests open the launcher and compare the
address bar, which stays mounted across the toggle.
Archive moved the active file into .archive and left the index flush to its
caller, so a failed flush kept the fact archived while MEMORY.md still named it,
with no way for a retry by stable id to tell which state the store was in.
The move is now undone when the flush fails, and a retry by stable id finds the
archived copy and drops the stale line. The test seam this needs is the one the
rename fix already added to the mainline.
RestoreArchived wrote the active file and reindexed it before removing the
archived copy, so a failure in either later step left the same id active and
archived at once, and the identity guard refused every retry.
Both steps now undo the restored file, so a failed recovery never leaves a double
identity and the same call succeeds once the obstruction is gone. If the undo's
own index write fails, a line naming the fact can remain; the retry rewrites it.
The quiet phone composer hid the branch pop along with the refine entry,
so the no-repository chip, which explains why git features are missing,
was unreachable at 390px and the branch-tooltip browser guard failed on
studio. Only the refine entry is hidden now; the composer guard asserts
the branch pop stays.
The perf page imported fewer stylesheets than main.tsx, so controls such
as the changed-files pill and the session row actions rendered unstyled
in anything captured from it.
The sidebar no longer slices each project to a fixed 30 rows: it asks the
kernel for pages and draws every loaded row. Batch selection now takes
all displayed rows, the selection pruning moves into a hook beside the
toolbar so Workspaces.tsx stays under its ceiling, and the two tests that
assumed the old 30-row cut assert the loaded rows instead.
A folded thread hid the maintainer's question while the status said
the report needed an answer. A report with needsInput now starts open
and is acknowledged as seen like any opened thread. The existing
entry-badge test opens the threads it expects to be read, and the
needs-input test no longer clicks a thread that is already open.
The six rules move out of studio.css into phonecomposer.css, without
the two !important the budget does not have room for: the attach
button's display no longer forces itself, so the quiet-state hiding wins
by specificity. The quiet toolbar also stops wrapping its model group.
The composer guard clears the draft left by its earlier steps before the
phone section and the guard page imports the new stylesheet.
The toggle was disabled whenever the settings sheet was open, so after
pressing any icon in the rail the top-left button looked live and did
nothing. It now opens and closes the workspace column in settings too.
The streaming summary line wrapped to a different height on every token
while the transcript was pinned to the bottom, so the card and everything
above it moved up and down for the length of the fold. The line is now
always drawn, one line high, truncated with an ellipsis.
The conversation, run analysis and workbench tabs sat on their own lighter
band with a divider under it. The row now takes the page's tone and drops the
divider; with a wallpaper the existing wallpaper rule still gives it a stable
surface.
Switching away from a built-in browser tab and back drew a white page until the
window was resized: a view parked offscreen and moved back with setBounds alone
keeps a widget that does not repaint. The shell already knows that attaching a
view again wakes its widget (it does so on minimize and restore); a page that
was parked now gets the same when it is shown.
#12340 added 18 lines to studio.css, which must not grow. The drop-hint rules
move to styles/sidebardrop.css, imported with the other feature sheets, and
the sidebar-drop fixture gains the Sidebar props that became required after the
branch point (navShown, paging) so the file typechecks again.
#12318 added 26 lines to studio.css, which must not grow. The quick-action
rules move to styles/sessionquick.css, imported with the other feature
sheets, and the paging test finds its filter chips inside the filter group so
the new archive and delete buttons, whose names also start with 归档, are not
mistaken for them.
Workspaces.tsx held the session row's menu inline, 63 lines of JSX that only
report which action was chosen. It is its own component now, with the same
markup and the same handlers, so the sidebar file has room for the quick
actions, folder drop and batch selection that are waiting to land in it.
SaveWithOptions wrote the new active file, then removed the old one, then
repaired the index. A failure in either later step left one id readable as two
active facts or left MEMORY.md naming a file that no longer existed, and a retry
could not tell which state it was in.
The rename path now stages the new content, retires the old name out of the
active namespace, flushes the old index line, publishes the staged file, and
reindexes; every step before publication is undone from the retired copy. The
helper that removed the old file in place has no callers left and is deleted.
Every writer that declared no extent (bash, MCP, a sub-agent given no
write_paths) held the whole workspace, so two sessions on one project queued
behind each other even when nothing they wrote overlapped. [agent] write_lease
= "optimistic" in the user file makes such a writer claim nothing; a writer
that declared paths, including paths that cannot be resolved, still excludes
other sessions. The mode is carried by each session's lease owner, so sessions
in one process keep their own, and a project file cannot change it.
Studio had no click entry for branching: /branch copied only the current tip and
swapped the pane onto it. A finished reply now carries a 'Branch into a new
conversation' button that copies the conversation up to the end of that reply's
turn into a new session, titled '<title> - branch' so it never reads as a second
copy of its source, and opens it while the original stays as it was.
Controller.BranchAt cuts at a live turn boundary and shares the file-and-meta
writing with Branch. POST /branch/at answers with the new path and refuses with
typed codes (running, unavailable, failed).
The catalog can now pull a published release, but a user already running it
saw an ordinary update. The version hub carries a withdrawn note for the
running version, mirrored on the desktop side, and the version panel leads
with it: which version, the publisher's reason, and where to move to.
A report that needs a log (it will not start, the kernel keeps restarting) had
no way to say where the log is: Studio writes shell.log and host.log under the
user data folder but nothing opened it, and the Windows and Linux builds have
no menu bar to put an entry in. The feedback form now offers 'Open logs
folder' when the shell can open it; the shell names the folder, the page does
not.
A 2000-turn conversation drew 2000 buttons per rail, 0.27 px apart, and every
pointer move rewrote all of them. Marks closer than 2 px are not told apart by
eye or pointer, so only one of each cluster is drawn; the rest stay addressable
through their positions, so the pointer and the keys still land on every
message.
The first screen fetched the 7.8 MB full Noto Studio font because every Chinese
character resolved to it. A core face with the common simplified-Chinese set,
Latin, punctuation and symbols now comes first, and the full font stays as a
later family limited to CJK ranges, so a character the core lacks still finds
its glyph there and only then pays for the download.
TestGoalResumeCommandContinuesTheGoalAndSaysSo failed in the race job on
studio (run 38188087939) with "resuming the goal said nothing" in 0.00s,
while the turn ran and the goal completed. The test built the expected
notice from c.Goal() as it read each event, and completing a goal clears
its text; the buffered events channel lets the controller send the notice,
finish the turn and complete the goal before the test reads anything, so
the real notice was compared with one built from an empty goal.
The expected line is now built once, before /goal resume is submitted.
TestResolveJumpHostsUsesConfiguredAliasesAndSSHConfig bound two things:
that ResolveJumpHosts reads ssh_config aliases and configured hosts, and
that this machine's real `ssh -G` answers. The second failed on Windows CI
again, twice on studio (runs 38179453323 and 38193303449), both at 15.0 s
with `ssh -G "bastion": exit status 1`: the product's own subprocess
timeout. The test's comment asked to split the two if it recurred rather
than shorten that timeout.
The resolution test now supplies the `ssh -G` answer through the source's
resolveOpenSSH seam, as the sshconfig tests already do, and keeps every
assertion. TestResolveJumpHostsThroughTheInstalledOpenSSH runs the same
alias through the real client, so a starved ssh fails under its own name.
Reviewing #12562 left three points open in this code.
differentialPerTest checked the base side before registering the defer that
restores the source, so a failure from that check returned with the tree still on
base - the state the restore exists to prevent. The defer is now registered
before the check, and treeMatches is a variable so a test can inject the failure;
TestDifferentialPerTestRestoresWhenTheBaseCheckFails asserts the PR's content is
back after that failure, and deleting the reorder makes it fail.
The report printed one sentence for every analysis failure, claiming the pin and
mutation figures were both absent even when the pin table had already been
produced and only the mutation sweep failed. analysisFailureLine now says which
figures are missing, and a unit test pins both cases.
The comment above the restore helpers described what the code used to do; it now
states only the constraint it enforces.
1.x closed an open slash or @ menu on Esc and, while a turn ran, cancelled
the turn as well. 2.x's menuKey consumed the key after closing the menu, so
a user who had started typing a follow-up during a run had to press Esc a
second time to stop it, while the thinking line said Esc stops it.
menuKey now cancels the running turn after closing the menu, except in vi
mode, where Esc never interrupts a stream. Idle, Esc on the menu still only
closes it and keeps the draft.
The sponsor logos link out, but the paragraph beside them read as plain text,
so the entry point was easy to miss. Each paragraph now ends with a visible
link to the same registration address its logo uses.
A release found broken after publication could only be fixed by shipping a
newer one, and installs kept being offered it meanwhile. The catalog now
carries a withdrawn list: a withdrawn version leaves the list old clients
read, a new client never offers or installs it and names it when it is the
one running, and a rerun of its own publication cannot list it again.
scripts/withdraw-version.sh edits the catalog; a manual workflow runs it
against the Studio catalog and keeps the GitHub release off the latest slot.
A tape is recorded as three files per provider exchange, and every write
discarded its error. A recording that lost a file therefore looked recorded, and
replay could not tell the difference: with the recorded request missing,
requestDiff compared against nothing and reported a divergence, blaming the
harness for a tape that was never complete. Nothing in the run's own report
mentioned it either, so a run whose tape can never be replayed read as a success.
save now returns its error and writes request.json last, so that file's presence
marks an exchange complete; load reports a missing recorded request as an
incomplete tape instead of a divergence; and the meter counts failed writes and
names the first error, which the report prints.
Two review findings in the same boundary are fixed here too. The run path read
the meter through record, which snapshotted before the final handler's deferred
save could run; stopping the meter is what waits for those handlers, so record
now drains first and the last exchange cannot be missing from the report. And a
re-recording over an existing tape left that tape's completeness markers in
place, so a failure partway through read as an old request beside a new partial
response; recording now clears those markers before it starts.
The healthy path is pinned as well, so the failure cases are not the only ones
covered: a complete recording reports nothing incomplete and prints no line.
Deleting any of the four checks reproduces the old behaviour.
A batch printed its own provenance to the terminal and kept the per-run records
on disk, so the two could not be read together afterwards. The fingerprint
(model, endpoint, window, generations, commit, start time) existed only in
scrollback, and every aggregate the report is read for — funnels, dose response,
pooled deltas — was printed rather than stored. Reconstructing a batch after the
fact, or attributing a figure in the README to the runs that produced it, meant
trusting a terminal transcript; the 24-run and 36-run totals in the last batch
are exactly that failure.
Each paying mode now writes batch.json beside results.jsonl: mode, start and
finish, the checkout's commit and dirty state, model, window, generations, the
task ids and arms the batch covered, planned against completed runs, the summed
usage of the runs that reported any, and the sha256 of both artifacts so a
manifest and the records it describes cannot drift apart.
Two things stay honest. Usage totals are absent when nothing reported any,
because an unmeasured cost must not read as a measured zero, while
usage_reported_runs is written even when it is zero, because that is a fact. And
a checkout git cannot describe is recorded as an unknown commit with no dirty
flag at all, rather than as a clean tree.
The icon rail kept its own icon set, so the same destination showed one mark
on the rail and another on the page it opens. Entries that open a section now
read that section's mark from the settings table; only the conversation item
keeps its own.
A history can carry a tool call whose arguments were recorded as an empty string
(a no-argument tool on a gateway that sends ""). Replayed as-is, a strict
gateway that parses every function_call.arguments as JSON refuses the whole
request with HTTP 400 ("Assistant tool call arguments must be valid JSON").
The repair that already closes truncated arguments now also turns empty or blank
arguments into {} on the way to a provider; a saved session keeps what it
recorded. The repair moves to its own file.
The baseline had diverged from studio's. Took studio's baseline.json and
re-ran repolint -update, which only lowered budgets for the code this change
removes (70 lowered, 14 entries removed, none raised).
reasonix upgrade overwrote the prebuilt binary inside node_modules, so
npm kept recording the old version (npm outdated -g, npm ls) while the
binary reported the new one. An executable that lives in an npm package
(bin/ next to a package.json naming reasonix or @reasonix/cli-*) is now
refused before any download, with the npm command to run instead.
The no-op main builds on every platform but Linux so Windows no longer has an
empty package, fixes its typo, and drops two comments that still described the
Windows helper.
Studio on Windows runs the downloaded installer directly (Line.RunInstaller);
nothing builds or starts reasonix-update-helper.exe there. Delete its Windows
sources and tests, the WindowsHandoff staging code, the Windows payload
manifest and its sign subcommand, and move the repair/installlayout entry
points that only that helper called into test support.
The process-input gate counts a direct os.UserHomeDir call; folder grants now use
the package's injectable osUserHomeDir like the other home lookups here.
A Build handed Options.Home still put three storage roots under the
process's home: isolated and best-of-n worktrees, browser profiles, and
the missing-reasoning recovery state. Each now resolves through the
build's config.Roots, which is the process binding whenever no home is
stated, as in every production frontend.
The new process-input repolint rule ratchets direct reads of the process
environment, working directory and user directories below
internal/frontend, so later slices can only move that count down.
40 boot tests now state their home and workspace instead of mutating the
process, and run in parallel. The isolation effect test runs two
sessions with different homes side by side and fails without the
binding fix.
The provider registry now holds a read-write lock: parallel Builds
register and resolve kinds concurrently, which -race reported as a
race and could crash with a concurrent map write.
Refs #12531
Refs #12521
The path is a manifest inside the source the person asked to install, and the
read is capped at the fetch limit. The CodeQL alert on the open is a false
positive for this call; say so where the open is.
The rebase onto studio put the unnamed-input surface three over its ceiling and
left provider.preset unregistered. The search, name and key inputs now carry the
provider.draft action with a value, and the gallery and install form buttons are
registered.
POST /providers/presets/install builds the entries a preset declares, whole
(catalog, windows, vision models, thinking shape, PresetID and version) with the
person's name and key, through the same storeProviderEntries the manual add
uses; the add form lists the kernel's presets grouped by host, searchable with
the form's own input style, and installs through that call.
The model cannot read the host clock, so a session left overnight keeps
reasoning from the date it started on. The date (with weekday and UTC offset)
is now a turn block owed through a projectionDebt: delivered on the first real
turn, absent on later turns of the same day so their bytes are identical,
re-owed on the first turn after the date changes and after a fold. It never
touches the cache-stable prefix. The clock is an explicit input
(control.Options.Clock); the boot assembly supplies the system clock and a
controller built without one tells the model nothing.
/tree sent "sessions": null for a workspace with no saved session, because
workspacePage returned a nil slice both when the listing failed and when
there was nothing to list. Studio iterates that field, so a fresh install
or a new project opened to a blank window.
Fixes#12743
RemoveProvider now returns which roles it handed to another service and which it
emptied, and also removes the name from desktop.provider_access. The remove
endpoint carries that report and the provider detail pane shows it, so a
removal that switches vision, advisor or decision off says so. A failure after
the removal was already written is shown where the service used to be instead of
in the detail of whichever service is selected next.
A job the session started and has not collected writes into the tree
without any receipt, so prose after a passing check was waived as if no
writer could still change files. The contract now counts a running job of
the turn's session beside tool hooks and extension sidecars.
The generic check was measured from the latest surviving write, so a
README or CHANGELOG edit after the tests passed sent the turn back to
re-run them. The anchor now steps over a trailing run of prose whose
extent the host established, when no check is declared and the role is
not Delivery, and only while a check stands passing over both the latest
change beyond prose and the anchor the turn would have had without the
prose. The prose judgement is the one #12520 introduced, now read per
receipt by both the whole-ledger waiver and this one.
A named write follows links, so its receipt now also names the file a
link resolved to; otherwise writing code through a .md link and removing
the link before the end of the turn would read as prose.
A configured tool hook, or an extension sidecar on a point every tool
call passes (tool.before, permission.decision, tool.after), runs outside
the bash sandbox and names no paths, so the check contract carries that
fact and neither prose waiver applies while one is present.
Provider clients took netclient.NewHTTPClient, so a redirect from the
configured address was followed by Go's default policy and the key went with
it. Add redirectguard.StayOnOrigin, build the probe, provider, System One,
balance and doctor clients through endpointclient.New, and guard the default
clients of the model listing, balance and System One calls. A refused redirect
is a permanent error that names the target host and reaches the user as
redirect_refused instead of a network fault retried ten times.
The test relied on a freshly written file's modification time being inside the
20ms window of the process clock. A file system stamps from its own clock, which
on a loaded Windows runner can lag the process clock by more than the window, so
the object was served from memory and the leg failed intermittently. The store
clock and the file time are now both set, which is what the test is about.
Merge Studio 0ac9826a9e while retaining Pane-owned undo and upstream work folding, running-tail handling and repeated-speaker grouping. Adapt three imported transcript fixtures to the existing undo ownership contract and cover folding, attachments and restore/undo together through the real Pane.
Ordinary upstream integration: 238 files, 7968 insertions and 704 deletions. Manual resolution and regression delta: 41 insertions and 9 deletions. Both histories are preserved.
Windows refuses < and > in a file name, so TestTreeWriteKeepsJSONContentTypeAndEscaping
failed at mkdir on the Windows leg. & is the other character the JSON encoder
escapes, so the same assertion holds there.
A move copies the root while the app keeps running and writing to it. The
verification pass hashed every file with no progress report, failed the whole
move on the first file that had changed since it was copied, and the commit
then deleted the old folder even when the app had written to it since.
Verification now reports bytes and files as it goes, copies again what changed
(bounded rounds, then ErrSourceChanging), and the old folder is kept when it
was written to after the proof. A refused target names what it holds that the
data does not own.
Replaces the stack #12601 (hide by default), #12637 (hint line) and #12643
(kernel setting). [ui] show_receipt (default off) is read and written through the
controller (ReceiptSettings, GET/POST /receipt-settings) like show_changed_files;
Studio keeps a per-kernel answer and carries an old browser-local on across once;
the TUI draws a line, or the card when the key asks for it. The kernel grades the
receipt (Severity) and the hidden card leaves one line for a graded one.
#12655 added one Messages field and pushed the file to 801 lines, which fails
repolint on the baseline. Drop one separator line between two field groups.
The column opened with a browser tab already in it, whichever way it was
asked for. It now opens on a start page that offers the file explorer, the
changed files and the browser, and a browser exists only once it is chosen.
A page a followed link already has on its way is still shown as before.
The sidebar had three toggles in two screen positions (header button,
title-bar button that shifted with the sidebar, edge grip), a rail
conversation item that did nothing from the conversation view, quick entries
repeated in the sidebar while the icon rail held them, and an R mark plus a
name on every reply. The title-bar button is now fixed to the window and is
the only button, the header one is gone; the conversation item opens and
closes the workspace column and returns home from settings; the sidebar's
quick entries show only when the rail is not drawn and the rail gains the
files entry; replies from one model after the first carry no header.
[ui] fold_turn_work (default false) is a kernel setting read through the
controller like the changed-files pill. When it is on, a finished turn
shows its last sentence and gathers the tool calls, thinking and the
sentences in between into one closed line with the call count and any
failure; the line opens on click and keeps its state, and the turn that is
still running is never folded.
At a couple of hundred CSS pixels the composer's model group could neither
shrink nor wrap and the title bar's right-hand controls left its title no
room, so rows ran over each other. Below 300px of room (the width at which
the five controls leave the title under its 4em floor) the bar gathers phone,
find, theme, sandbox and browser under one button; each stays one press away.
The model group wraps, the tabs shrink, the meter row drops its cache figure
and the placeholder keeps to one line at the narrowest widths. The size
slider's fallback range parks the thumb mid-track while no range is announced.
The run analysis drew a rail labelled only 'Signal rail' and 'Shares the
timeline below'. A line now says it holds the run's records that have no
duration and are not tool calls, placed at the moment they happened, so the
position is the time and cannot be dragged.
Choosing a folder in the system dialog no longer grants it; Add does. The
panel states that the agent can change and delete anything in an added
folder and that panes already open narrow only on rebuild. A failed read
shows a retry, refusals use the error box, and the project row is laid
out like the other facts.
Settings > Tools and permissions > Sandbox lists the project folder and
the extra folders granted to it, and adds or removes them through the
existing /workspace/roots routes. The kernel's refusals are worded by
their codes. The panel says that 'allow everything' only decides whether
to ask and does not widen where writes can land. No folder is added by
default.
[notifications] sound = true plays a short system chime for the events already
selected, independent of system notifications and of their muting, so a
minimised or backgrounded window is still audible. Off unless asked for. The
Windows, macOS and Linux senders implement Chime; the Settings notifications
block gets its own switch.
Import Studio 71650fe7, including upstream desktop lint fix 85460ae and test-only sandbox portability fix 2b5f6448. Keep the running-jobs fixture deterministic by selecting its scripted provider explicitly and asserting the background launch.
Integrate Studio 71650fe7, including upstream desktop modernization 85460ae14 and the job-list test convention from 2b5f6448. Keep the latter test-only setting unchanged and use the supported scripted resolver with explicit selection and launch assertions. The only authored delta from the automatic merge is 17 additions and 10 deletions in that fixture; no coverage workflow or product sandbox changes.
In the activity group a running call that had printed nothing was a row that
did nothing when clicked, and a long command was clipped to one line, so a
call waiting out a timeout looked frozen. Such a row now opens to its input,
says it has no output yet and for how long it has run, names the command's
own timeout when it carries one, and its hover text holds the whole command.
Attachments, wallpapers and theme packs were each turned into base64 before the
kernel could refuse them. One guard in the port checks the kernel's cap first,
using the image cap for pictures, and a Go test pins every number to the kernel's.
The test starts a background bash job through the real Build assembly. On a
host without bubblewrap the sandbox refuses to run unconfined, so the job
never starts and the coverage leg fails there while passing on a developer
machine. Its sibling effect tests already declare [sandbox] bash = "off".
Windows 11 kept the kernel's 10.0 major, so a build like 10.0.26200 read
as Windows 10 (issue #12696). Builds from 22000 report as 11.0.build,
matching the product version the darwin path already reports.
/tree windows each workspace by the request's scope (active, archived) and word,
always builds pinned active conversations, and reports the total, archived and
remaining counts for the whole folder. The sidebar counts from the kernel, so the
chips do not move with the window, an archived conversation far outside it is
reachable, search is filtered on the kernel and paged, and a folder holds at most
300 rows. The remote machine list shows the far kernel's total instead of a
truncated count.
Windows accepted \??\C:\... because only a leading double backslash was
refused. A root's volume must now be a drive letter and a colon; UNC, \\?\, \\.\
and NT object paths are all refused. Adds a boot-level test where a delegated
child writes only a granted folder.
PowerShell candidates were launch-probed one after another with a 10 s
timeout each, and nothing bounded host construction or the lookups around
the probes. On a machine where Store aliases, a PATH entry or a registry
read stall, first start sat in shell discovery past the shell's 60 s
handshake wait.
PowerShell now shares bash's hedged search and budget, probe commands
cap their post-kill wait, and the whole resolution (host construction
included) gives up after both budgets plus a hedge: it uses a bash
already proven in this process, else the bare name with a warning that a
restart usually recovers.
The skill-shadowing warning after an install and the not-selected warning
after a root is registered now render their paths through hostLiteral
(60f37ee7c), checked by a unit test that feeds an unterminated OSC to the
two sentence helpers. #12547 also asked for the newline, bidi and
zero-width cases.
Two end-to-end tests add them: a skill is installed, or a root
registered, while another skill under a configured path whose name holds
the character keeps the selection, and the warning must show the
character escaped for every path rather than carry it raw. The newline
case is skipped on Windows, which cannot hold one in a file name, and
runs on the Linux and macOS jobs.
Fixes#12547
Each transcript row asked the browser to skip rendering it while off screen
with an estimated height, and swapped the estimate for the measured height
the first time the row scrolled into view. The page shrank under the wheel
(12975px to 11217px in the probe), so the content jumped, once by 1858px.
Rows that are far off screen are already unmounted by their block, which
holds the measured height, so the estimate had no job left.
unapplied_steer names the queue entry the kernel took out of the turn
without the model reading it and holds for a retry. The row that said it
was waiting stayed forever, while the queue panel, which the kernel
authors, showed the same line as blocked and retryable. The row now goes
when that notice names it; the notice card stays.
The 400 text told every reader it was probably a bug. The host cannot know that:
the reason the service gave is appended, so the preface now says whose refusal
it is, where the reason is, and what can be tried, without promising it works.
The poll re-ran setContextMenu every five seconds with an identical menu.
Where the menu is handed to the desktop's indicator over D-Bus, replacing
it closes one the person has open. It is now rebuilt only when the
sentence, the words or the close-to-tray setting differ.
The prose waiver judged paths by what the disk showed at the end of the turn,
so a documentation-named symlink to a code file that was written through and
then removed counted as a deleted prose file. The resolved target is now
recorded on the write receipt while the link still resolves.
Refs #12524
attach() turned the whole file into an ArrayBuffer, a binary string, base64 and a
JSON string before the kernel could refuse it at 25 MB, so a modpack-sized
archive could take the renderer down. The size is checked first and refused with
the kernel's own attachment.too_large identity.
The rail's hints were drawn the instant the pointer touched an icon, so a
pass across the column flashed every label. They now appear after a 0.3s
rest (--tip-delay; VS Code waits 250ms), and keyboard focus still shows its
hint at once.
The [desktop] renderer wrote conversation_width only for full, so a full-all
choice reverted to standard on the next start. It now writes every non-standard
width; a render-and-reparse test, a page-attribute test and narrow-viewport
checks cover the new choice.
The font fields took any family the machine has but suggested only a dozen,
and on Windows few of those are installed. The suggestions now cover the
common CJK, Latin and programming faces, filtered to the ones the machine
really has, as before. Nothing is chosen by default.
The full measure widens the transcript only, by design. A third choice,
full-all, moves the composer's column with it; the default and the existing
full measure are unchanged.
A host-authored inbox item (a background job's completion, for one) ran
as a synthetic turn that opens no boundary and so emitted answer text and
a turn end with no turn start. A client that never heard the start kept
the send arrow, showed no stop button and counted nothing running, and
its status poll stayed off, so it could not stop a turn that was
running. The queued host turn now announces its own start.
A refusal reached the model only in that call's result, which compaction can
fold away, and the posture refusal did not say not to retry. Calls refused by
the host (any permission refusal code) are now owed to the next real turn as a
refused-calls block, delivered once like the other turn projections, and the
posture refusal says the call and its rewordings will not be approved.
Refs #12253
On a 1M window the fixed 8K output margin was smaller than the prompt
estimate's own error, so a prompt a percent above its estimate plus a 64K
completion exceeded the window and the provider refused every turn. The margin
is now max(8K, window/32), and a codeless 400/413 is classified as an overflow
when the estimate plus the requested completion is past the window, so the
turn folds and replays instead of repeating the refusal.
Fixes#12694
Windows 11 reports NT 10.0, so the environment block read like Windows 10.
The form now shows 'Windows 11 (10.0.26200)' for builds from 22000; the
version string that is sent is unchanged.
The kernel sent toasts through the Windows Runtime but never installed an
activation callback, so a click did nothing. It now registers one; the
host forwards it as a show act over the handshake pipe and the shell's
act handler (now its own module) shows and focuses the window.
[agent] ask_timeout_seconds (user file only, default 0 = wait) bounds how long the
agent's own ask question waits. On expiry the host settles the card on every
frontend with a timed_out decision receipt and returns the model a result that
names the timeout, not an answer. External forms are never skipped.
Studio settings and the terminal's /ask-timeout drive the same controller
setting.
Extra roots are the project grant's folders in the user's home. They are a set
of their own: write roots and the OS jail read them, the guard on Reasonix's
own state does not treat them as exemptions, the turn tail names them, and the
scan, readiness and receipt of the executor and of every sub-agent count them
as workspace. One judgement (FolderGrantRefusal, shared with workspace trust
and POST /sandbox) refuses drive roots, home and its ancestors, storage roots,
key and credential directories and the forbid list, in either direction,
comparing file identity rather than spelling, plus folders the user does not
own, UNC and device-namespace paths, and overlaps with other write roots. Every
load vets each grant again. Withdrawing a root rebuilds every pane on the
workspace, after this pane's lock is released.
The change time cannot be restored the way a modification time can, so an
in-place edit that restores size and mtime misses the cache. Platforms
without that identity fall back to size and time. Adds limit and
replacement tests and states the in-process memory in the package doc.
A locked branch-meta read-modify-write passes what it read to the write
instead of reading again, and recording a revision no longer re-reads
what it just wrote. trustedstate.Object remembers a verified object with
its size and modification time and serves it while both are unchanged
and old enough to be meaningful; Verify always reads the bytes.
The docked pane's grid kept the navigation and find rows across both
columns, so the browser column began 44px below the pane top and its own
8px padding made the gap 52px. The column now spans every row beside them;
under the scene fold the old stacked layout is kept explicitly.
stageDispatcher built the successor with empty options, so after a subgraph
rebuild an optional extension without a sidecar was skipped silently, right at
the moment an install or enable makes that most likely. The successor now takes
the reporting of the dispatcher it replaces.
Refs #12258
Bumps the go group with 2 updates: [github.com/mattn/go-runewidth](https://github.com/mattn/go-runewidth) and [golang.org/x/sys](https://github.com/golang/sys).
Updates `github.com/mattn/go-runewidth` from 0.0.30 to 0.0.31
- [Commits](https://github.com/mattn/go-runewidth/compare/v0.0.30...v0.0.31)
Updates `golang.org/x/sys` from 0.48.0 to 0.49.0
- [Commits](https://github.com/golang/sys/compare/v0.48.0...v0.49.0)
---
updated-dependencies:
- dependency-name: github.com/mattn/go-runewidth
dependency-version: 0.0.31
dependency-type: direct:production
update-type: version-update:semver-patch
dependency-group: go
- dependency-name: golang.org/x/sys
dependency-version: 0.49.0
dependency-type: direct:production
update-type: version-update:semver-minor
dependency-group: go
...
Signed-off-by: dependabot[bot] <[email protected]>
Keep the already-built quote and broadcaster projections when canonical configuration cannot be read. Warn without rewriting the recoverable file, while still enrolling valid local controllers. Cover real Adopt, Open, model, extension and workspace publication with malformed configuration, including built-in-default fallback. Refs #12674.
Assisted-by: OpenAI Codex
Bumps the go group in /desktop with 1 update: [golang.org/x/sys](https://github.com/golang/sys).
Updates `golang.org/x/sys` from 0.48.0 to 0.49.0
- [Commits](https://github.com/golang/sys/compare/v0.48.0...v0.49.0)
---
updated-dependencies:
- dependency-name: golang.org/x/sys
dependency-version: 0.49.0
dependency-type: direct:production
update-type: version-update:semver-minor
dependency-group: go
...
Signed-off-by: dependabot[bot] <[email protected]>
Preserve the controller retirement identity at the HTTP boundary with a localized 409 refusal. Exercise selection racing a real model replacement, then delegate to the retired controller and assert no config or currency mutation. Refs #12674.
Assisted-by: OpenAI Codex
Exercise real booted local runtimes, a remote pane, live ledger and balance projections, future wire quotes, model replacement and sequential stale calls after close. The canonical-config fixtures do not claim a successful durable settings write. Refs #12674.
Assisted-by: OpenAI Codex
Verify canonical quiet enrollment, failed binding atomicity, replacement and close membership, and projection ordering against the next notice. Refs #12674.
Assisted-by: OpenAI Codex
Check full local sinks, future quotes, remote exclusion, retirement and failed-save atomicity. Retain the real HTTP currency-save assertion for CI; this sandbox cannot acquire the OS-user config lock. Refs #12674.
Assisted-by: OpenAI Codex
Enroll only local controllers and reconcile their quote and broadcaster projections before publication. Share the same ordered replacement path across model, extension and workspace rebuilds, and retire membership when a runtime closes. Refs #12674.
Assisted-by: OpenAI Codex
Serialize persisted display-currency changes with local-controller membership. Quiet enrollment refreshes the exact quote context from canonical configuration; retired generations cannot save, and successful saves notify each full local sink. Refs #12674.
Assisted-by: OpenAI Codex
Bind unread polling to availability and identity of the first local machine owner. Opening or replacing that owner starts a fresh read while ordinary active-pane switches keep the polling session. Cover remote-only startup, opening local sessions and retiring the first local owner through the real App. Refs #12674.
Assisted-by: OpenAI Codex
Use the same pinned local catalogue owner for the default-model write and read. Cover a model exposed only by the second local runtime through the actual Settings picker, SSE route and HTTP catalogue validation. Retain the durable-write assertion for CI; this sandbox cannot acquire the OS-user configuration lock. Refs #12674.
Assisted-by: OpenAI Codex
Pull requests whose diff is provably platform-neutral run only the Linux legs;
the macOS and Windows legs keep their check names but execute on an ubuntu
runner with every step skipped. Pushes to main-v2/studio are unchanged.
Give the benchmark window one stable BenchPort owner so its language, zoom, theme and save instrumentation still reach usePaint. Exercise the real harness and paint hook without changing browser guard assertions. Refs #12674.
Assisted-by: OpenAI Codex
Select the active local pane, else the first local pane, for Memory, model catalogue resolution and legacy-session recovery. Keep machinewide pages on their stable host owner. Retire cached workspace settings ports after successful close and restore the unrelated portFor comment. Refs #12674.
Assisted-by: OpenAI Codex
Allow host settings requests to pin an existing local runtime. Refuse stale or remote owners without falling back, while unpinned machine settings retain the first live local owner. Cover real workspace memory mutations, model routing and runtime replacement. Refs #12674.
Assisted-by: OpenAI Codex
The copy claimed alerts appear while the window is in the background; the sender
fires whether or not the window is in front. It now says so, says the switch
ships off, and that the sound is the system's with no separate setting.
The update path started the NSIS installer with no arguments, so it ran
as a fresh install: the previous uninstaller deleted the shortcuts and
removed the AppUserModelID entries (the taskbar pin), and the installer
recreated the desktop shortcut. With --updated the installer takes
electron-builder's in-place update path and keeps them.
Add a default-off Appearance preference for session tabs, persist the choice locally, and show the existing strip only when multiple session panes are open. Preserve single-session behavior and cover preference changes, persistence, and unread tab marks with interaction tests.
The orchestrating agent had no way to ask which background jobs are still
running and for how long, so it kept a count of its own and miscounted. The
listing comes from the job manager that owns the jobs (id, kind, label,
elapsed), through the existing read-only tool so no tool schema is added.
list_subagents rows now report when a child started and, while it runs, for
how long.
The queued line is drawn after the message the model is still writing, and
appending the next chunk only looked at the last item, so the chunk opened a
new message and one stretch of thinking showed as two blocks. Appending now
continues the open message past lines that are still waiting; once a line has
been delivered and the earlier message settled, the next chunk is a new one.
The branch chip's small "N changes" moves to a centred pill right above the
input box reading "N files changed +x -y", from the same tree.changes the
workbench lists. Clicking it opens the workbench's change section. A kernel
setting [ui] show_changed_files (default on) switches it off, which restores
the chip's small count.
TurnDone carries the run time up to the final answer; history carries it on
the answer that closed a turn. The terminal UI and Studio draw it under the
answer, and the live status timer uses the same compact form.
The Linux helper's accepted package name was a build-time variable that
linux-payload.js never set, so the helper in the .deb kept
reasonix-desktop and refused reasonix-studio as a different package.
It now takes the name from update.StudioLine, the one place the shell
asks dpkg by, so there is no build value to forget. A package command
prints it, studio.yml asserts the packaged helper accepts the .deb's own
package name, and a test ties electron-builder.yml's deb and rpm names
to it.
A read-only worker cut off by cancellation left a transcript the parent never
heard of: the aggregate said only context canceled. The reference now travels
with the cancellation (parallel_tasks, fleet, task) and read_subagent_result
renders what a cancelled child had said, called and read, labelled as not a
result and its tool output as untrusted. ReadFinalAnswer still refuses any
run that did not complete, so fleet adopt_ref cannot adopt one.
DELETE /themes/{id} removes a pack from the user's library only; the
listing marks those packs removable, and Appearance offers a two-step
delete on them. A bad id is a 400 theme.bad_id (ErrBadID), never a path
in the response. The config is settled before the files go and restored
if they fail, so it never names a pack that is gone.
The workbench handle lay full height over the transcript's scrollbar while
shut, and the open handles reached into the scrollbar on either side of a
seam, so a press on the thumb resized the column instead of scrolling. The
workspace handle now starts inside the seam, the docked transcript steps in
from the workbench seam, and the shut workbench handle is only as tall as its
grip, centred and set left of the scrollbar. The grip starts the drag itself
once the press has travelled.
/goal resume only flipped the state and said nothing, so the Goal sat idle until
another message arrived. It now says the Goal is resumed and runs the next turn
of it. A Goal the model or evaluator reported blocked also resumes on the user's
own next message, as one gated by await_user already did; the continuation epoch
is read after that lift so the lifted turn is advanced by the Goal instead of
being discarded as stale.
Refs #12554
The switch refused with busy.switch_model_jobs because closing the outgoing
controller cancelled its jobs. The replacement now adopts the job manager and
the workspace write lease, and the outgoing controller is retired without
closing them; its cleanup runs once the jobs it left have finished.
The admission notices for a session switch, a workspace switch and a draining
runtime told every frontend to resend, but Studio already queues the refused
line on the typed busy code, so following the advice sent it twice. The notice
now carries input_refused; Studio words it as queued, other frontends keep the
English text.
Refs #12554
The notes ceiling clipped the dormant-rules sentence and pushed the dismiss
button to a hidden third row. The button now shares the sentence's first row,
the ceiling has a floor outside short windows, and the queue yields the
difference. Also keeps the diagnostic for payload codes whose detail was not
consumed and pins malformed-payload cases.
Skill shadow and unselected warnings, the migrated and remediation
sentences (plugin name), the persist-status warning, manifest read
warnings and the strict-decode error now show package data through
ShownLocator or ShownIdentity.
Shadow warnings, the incompatible-plugin warning and the skipped
marketplace warning now write the whole sentence first and render the
package-supplied path, error text or reason with ShownLocator, so an
unterminated escape in that data cannot swallow the rest of the sentence.
conclude_blocked and await_user tell the model the turn ends, but a code write
with no check after it still produced a verification gap, so the host ran two
more rounds and reported the turn as failed. A successful hand-back after the
latest write now yields the continuation; the ledger still owes the change and
the completion receipt still reports it as unverified.
Refs #12554
Display, launch and stored policies share one scrub core, so the report
that a launch line lost characters is produced by the same pass that
deleted them (ZWJ, ZWNJ, bidi, escapes) and never by a second rule. The
marker leads the line so tail truncation in serve and the TUI cannot cut
it, and is worded through i18n. observe.Sanitize delegates to the stored
policy; the policy table is declared in doc.go.
TurnStarted and Steer events, and the rebuilt history, name the images under
.reasonix/attachments that a message references, resolved by the same
detectRefs pass that decides what the model sees and reported under their
normalized path. The transcript previews from that field instead of matching a
spelling in the text; the model still reads the @path text.
The runtime bar resolved variables for only two notice codes, so
permission_rules_dormant and job_* showed {n}/{list}/{rule}/{name} verbatim
plus the raw JSON payload. NoticeCard and RuntimeBar now share noticeWording();
a test fails when a code with placeholders has no registered reader.
The bash job function wraps the process's own ending in jobs.ExitError,
the failed-job notice payload carries exitCode and signal, and the
terminal UI and Studio word the sentence from them. The raw error text
stays on the payload and is what an untyped failure still shows.
Display (prose) now drops bidirectional and format characters exactly as
the prose policy does, where it used to keep bidi controls. The launch
line keeps deleting invisible characters, because redaction has to match
a key a zero-width character split, but it now says so with a marker
instead of reading as an unaltered line. One hostile fixture is asserted
across the locator, prose, display and launch policies.
The shell always opened at the default size, so a window the person had
sized, moved or maximised came back as if nothing had been chosen. The
normal bounds and the maximised flag are now recorded as the window settles
and as it closes, and the next launch reopens there. A rectangle no display
shows any more keeps its size and lets the platform pick the place.
The workbench handle lay full height over the transcript's scrollbar while
shut, and the open handles reached five to six pixels into the scrollbar on
either side of a seam, so a press on the thumb resized the column instead of
scrolling. Open handles now start one pixel inside the seam, a shut handle is
only as tall as its grip, and the grip starts the drag itself once the press
has travelled.
The message column shrinks to fit its content and a textarea at 100% of it
starts at its default width, so the rewrite box stayed about 270px wide in a
760px transcript. While a message is being rewritten its column now takes the
whole row.
The transport now returns ErrAuthenticationRequired (via AuthRequiredError with
server, resource and scope) for a 401 that carries a Bearer challenge on a
server without explicit credentials. InstallResultForError maps that identity
to action_required/authenticate and no longer matches words in the message or
promotes a bare 401/403 status. Adds a regression test that refresh stays at the
token endpoint stored at sign-in. The guide stops promising a Studio sign-in
control that does not exist yet.
The receipt took its success from the result after extensions rewrote it, so
a code write refused or errored by tool.after left no mutation in the ledger
and, as the only change of the turn, no verification was asked for. A named-path
write now keeps the success of the tool's own execution.
Refs #12528
The feedback timer spy counted testing-library's own waitFor deadline once
asyncUtilTimeout became 10s; it now ignores timers armed from inside the
library. The settings-layers case swallowed a lookup that never matched its
textbox name, which cost a full asyncUtilTimeout; it now asserts the real name.
vitest gets testTimeout/hookTimeout of 60s, and find/zoom-keys drop their
private copies of the same limits.
The shadow bundle inlined the accepted contract's criteria and the full
verdict obligation list on every turn. Both scale with the number of
captured tests, and the bundle bytes changed each turn, so the
content-addressed store kept a fresh multi-megabyte copy per turn.
The bundle now names the contract revision by record digest only, and
files the obligations as their own object, which repeats nothing when a
turn leaves them unchanged.
Messages is back under the line ceiling with room for the next notification kind.
A report whose last update is over thirty days old no longer keeps the unread
poll running.
Fedora, RHEL and openSUSE cannot install the .deb. electron-builder now
also builds an .rpm, signed and mirrored with the same minisign flow, and
studio-manifest lists it under native_packages[linux-<arch>-rpm] beside the
unchanged .deb key. fpm spells the architecture of an .rpm name x86_64; the
packaging hook renames it to the release architecture and fails on any
spelling it does not know. The package job reads the rpm back for its name,
the helper's owner and mode, and the rpm's post-uninstall leaves the command
link alone on upgrade.
The graphics panel no longer infers a driver fault from gpu_compositing: it
shows the raw status Electron reports, tells a saved choice from a launch-only
override, and says so when no status exists yet. The settings search moves into
its own hook so Settings.tsx has room again.
When the saved default_model is a decision source, a build that did not
ask to open on a fallback (reasonix run, the TUI) passes it over and
opens on the first conversation source, and said nothing about it:
ResolveNewSessionChatModel returned only (ref, fallback, ok), so the
boot had no skipped default to report, while the window path's
ResolveStartupChatModel carried one and emitted the notice.
Both resolvers now return one value, NewSessionModel{Ref, Fallback,
SkippedDefault}, and one helper decides that a saved default was passed
over: nothing configured serves it, or its source does not answer
conversation. newSessionModel takes SkippedDefault from whichever
resolver the build uses, so both paths emit the same
default_model_unavailable notice. What each path opens on is unchanged;
an unknown default is still kept on the no-fallback path for the CLI's
error, so it is not reported as skipped there.
Refs #12561
Record a digest of every tool definition the first time an authorized
server lists a non-empty catalog. A later listing that adds a tool or
changes any definition field is held out of the model's catalog and cache,
and a call to it is refused with the typed code mcp.tool_definition_held.
The user is told through a notice (mcp_tools_held), and accepts the held
definitions with 'reasonix mcp trust <name>' or POST /mcp/trust, which
advance the record and reconnect the server.
User-level servers keep one record for every workspace; project-level
servers keep one per workspace. Removing a server forgets its record.
Refs #12356
The announcement compared only the aggregate unread count, so another reply on a
report that was already unread was never announced. Each report's latest
maintainer reply id is now tracked as well.
When a .mcp.json is refused while planning (over the server limit, an
invalid entry), kind=auto hid the reason. A fetched .mcp.json fell
through to the skill attempt, and with strict:false the JSON was planned
as a skill named "mcp". A GitHub repository whose root .mcp.json was
refused ended as a remote HTTP MCP plan at github.com (a repo named like
"mcp") or as ErrUnsupportedKind. A local .mcp.json file ended in "no
installable ... found", and a folder carrying a refused .mcp.json
dropped it without a word.
A source that is itself a .mcp.json, fetched or local, now answers with
its refusal under kind=auto as it already did under kind=mcp. A folder
or GitHub repository whose .mcp.json is refused keeps planning its other
installables; a folder warns with the path and the cause, and when
nothing else plans, the refusal is the error. A .mcp.json that is absent
or declares no server says nothing, as before.
readMCPJSON no longer returns a bare filesystem error: an absent file is
ErrManifestMissing and one that cannot be read (a directory, a denied
read) is ErrSourceUnreadable, so the caller can tell the cause.
Refs #12514
An unread finish keeps a conversation out of the automatic sweep until it is
viewed. A hand-edited days value past the limit is held at the limit instead of
wrapping time.Duration. Pins are drawn from the tree's own pinned field;
localStorage only carries pins made before the kernel kept them and each is
sent once.
POST /submit answered 409 busy.session_running whenever the turn it had just
started was already over by the time the handler sampled Running(), and also
for a submission parked behind a finishing turn that would run next. The
submit entry points now return the controller's admission result and the
handler refuses only a submission that was dropped.
runTrace gave a fixture turn a fixed 15s to close. A turn is several rounds of
host work, which a loaded Windows runner stretches well past that while the
turn is still emitting events. A turn is now hung only after a silent window,
with a total ceiling for one that never stops.
Add a web app manifest, icons and the touch-icon and theme-color tags to the
page so a phone can install it and open it full screen. No service worker is
registered, so nothing cached can outlive a pairing or a credential. The serve
host now types .webmanifest as application/manifest+json.
The picker leads a row with an accent dot when the kernel reports it unread.
The TUI posts /sessions/viewed when it binds a conversation (launch, resume)
and when a turn ends in the one on screen while the terminal has focus; a turn
that ends behind another window is marked when focus returns.
First slice of treating a WSL distribution as a machine a kernel can run on.
Adds a read-only utility package that lists distributions from the structure of
wsl -l -v and checks one by its architecture and user, plus
reasonix remote wsl list/test. Nothing is installed or started.
An embedded struct's fields are its outer type's own on the wire, and an
interface's extended parents are part of its shape, but the checker read neither,
so the outer PermissionRules type could not be declared and a field added to it on
one side went unnoticed. Same-file embedded structs and extends parents are now
flattened; PermissionLists and PermissionRules are registered.
The first line of a snapshot, the app list and the screenshot caption printed
the application's name raw, and the app list printed window titles escaped
but unbounded. Route them through the literal the modal names already use.
A pin the kernel refused is rolled back and reported instead of staying
visible, and any remembered pin of a listed conversation the kernel does not
report as pinned is sent again on the next tree load, so a legacy pin in a
workspace added after the first sync is protected before its first sweep.
A re-delivered turn_started naming the turn already in front of us is the
same open twice. The transport folds the duplicate seq, but the reducer is
fed raw sequences, and sealing on order marked its live calls "no result"
and dropped an unanswered approval. The seal now keys the start's msgIndex
against the turn the live side holds; naming still runs on the duplicate,
because the row may have been sent between the two deliveries.
The record cannot end a turn. The kernel commits each text-only assistant
round before handleFinalResponse decides to continue it — baseline
criteria, executor-handoff nudge, empty-final retry, steer drain — and
/history carries no end marker, so rebuild stopped inferring "over" from a
record ending on a say. The typed fact rides every status read: Pane's
applyStatus dispatches __running into the reducer, equal answers fold to
the same state, and a payload that cannot say leaves the wire's running
standing. The identity merge stays; the record no longer decides the
turn's life. A lost turn_done self-corrects at the next status answer.
The rebase folds studio's extension retention into rebuild(): an
extension is not in the record either, so the tail keeps it the way the
old __restore filter did.
The status-inflight fixture now answers running over an emitted
turn_started: the pane reads that fact, and a port contradicting its own
wire is not a kernel these reads can be reasoned against.
A __restore re-read taken mid-turn — by hand, or by the stream-gap
rebuild — dropped the cards being written, letting the next delta mint
the answer anew: new identity, the entrance replayed over half a reply,
the text before the gap gone until the message frame. The tail now
merges with the record by identity instead of riding after it: the
kernel commits each assistant round at its boundary, before its calls
run, so the record can already hold a call the live side holds open —
the record wins where it is complete, the live card where it is not.
A record ending on the committed answer, with no call of ours still
out, is a turn_done the wire lost: the answer wins and the turn is
over, clearing a running flag nothing else would clear.
A turn whose end never arrived is sealed the way turn_done would have
sealed it when the next one starts, so its calls stop spinning through
a turn that has nothing to do with them.
session.ts stays under the 800-line ceiling: the merge lives in
state/rebuild.ts.
The read-only execution guard admitted only list and inspect, so
use_capability action=search fell through to the unknown-action block even
though the tool declares it and resolves it exactly like list. The tool's
own description tells every model to search before calling a tool
unavailable, so a read-only subagent was steered into the block.
usecap now exports the declared actions and the catalog-read subset; the
guard reads that set instead of keeping its own copy. The refusal for an
action outside the declared set names the action and the allowed ones.
Every boot.Rebuild re-derived each build-time notice (dormant permission
rules, held project MCP servers, default-model fallback, missing key) from
unchanged state and emitted it again. The runtime owner now carries a notice
ledger; each build runs through a window that drops notices the previous
successful build already showed, so only new or changed notices are emitted.
A Windows file name cannot hold a newline, so the quoting test could not create its directory. The bidi override still exercises the literal-quoting path there.
A removal refusal rendered in a banner at the top of the settings pane, out of
view once the pane scrolled, so Delete appeared to do nothing. The banner is
now an inline alert in the provider detail header. The list is re-read after a
failed removal too, because the kernel may have written the removal before the
failure. RemoveProvider now clears the guardian, recovery, vision, triage,
decision and advisor roles that named the removed service, walking the same
role table the other repairs use, and its doc comment states the behaviour the
code and tests already have.
A rule naming a tool that assembly registers only under some configuration or
in a sub-registry (system_one, advise, best_of_n, apply_isolated,
discard_isolated, submit_plan, conclude_no_changes, review_report,
complete_subtask) was refused on save and reported dormant at load, because the
tool is absent from the registry at that moment. Each owner now declares its
conditional tool with tool.RegisterConditional and the controller's vocabulary
reads the declarations; a boot test reconciles every tool the assembly can
register against the built-ins plus the declarations. The dormant-rule item
shape is also held by wire-parity.
The edit form showed 'read available models from provider' for a decision
source, whose protocol has no model listing. The provider entry view now
carries canListModels, derived from the protocol table's Discovery
declaration, and the editor hides both listing controls from it.
The authoring guards were three hand-written lists in corpus_test.go: which
corpora must carry a reference solution, which must have a grader that accepts
it, and which must let the agent decide when to stop. A seventh corpus could be
added and be checked by none of them, which is what happened to four of the
seven that exist.
Each corpus now declares, in one registry, how it stands against each rule:
enforced, or exempt with a reason. Discovery walks the corpus root, so an
unregistered corpus is a failure rather than an omission from a list. The three
guards read the registry instead of their own lists, and the stop-policy guard
covers every registered corpus rather than the three that happened to be listed.
An entry must name the corpus this root owns: its directory is compared with the
corpus's path under the root, not with its name, so a registry entry pointing at
some other directory that happens to share a name cannot stand in for the
committed one.
The pruning test the registry replaces was weakened rather than extended: a new
corpus used to be silently unchecked.
The model switcher listed every configured source, so a decision wire
(kind typesafe) appeared beside chat models and choosing one failed in the
provider registry as an unknown kind.
GET /models now answers one job: chat by default, answers=decision for the
decision sources, answers=all for the management views. The switch, the
default model and every role refuse a model whose wire does not answer that
job with a typed code (model.decision_only, model.not_decision_source), and
the provider builder refuses it with config.AnswersMismatchError. A saved
default that names a decision source is passed over like a stale one.
the contract translation (disabled -> none/low, enabled -> default) now
applies only when the menu is the contract, the same rule
InheritedEffortOfficial already used. Before, the Token Rhythm preset shape
sent thinking=enabled with reasoning_effort=max for a stored disabled.
Problem: PR #12348's browser menu dismissed agent-owned UI entries instead
of closing the underlying pages, overlapping the controller close work in
PR #12067.
Root cause: The tab strip owned a presentation-only close path separate
from the controller's browser lifecycle.
Fix: Integrate the latest studio base with #12067, retain its useTabClose
owner, and adapt #12348's WorkbenchTabs extraction, close-to-the-right menu
and keyboard access to that path. Keep close-all, preserve file editors,
and retain failed pages while disabling repeated closes in flight.
Verification: 99 focused frontend tests, TypeScript/production and perf
builds, census and its 10 fixture comparisons, make check, go vet, make
lint, builtin/boot and browser/serve/control tests, and 11 release-note
tests passed. Browser keyboard/menu acceptance and eight fixture screenshots
cover light/dark and wide/narrow layouts. Intrinsic, contrast and side
guards passed; wheel exits successfully with its two existing advisory
findings. Native Electron acceptance and hosted CI remain outstanding.
Co-authored-by: KHG420 <[email protected]>
Co-authored-by: par73e <[email protected]>
Both analyses edit a PR's source, run the tests, and put the source back. The
write-back discarded its error and then marked itself restored, so a failed
restore was indistinguishable from a successful one: the sweep went on to report
caught/survivor counts, and the differential went on to report pin results, for
a tree that was still mutated. A verdict from a tree that no longer holds the
PR's code is worse than no verdict, and the mutation would leak into the next
mutant and into the user's checkout.
restoreSource writes the bytes back and proves they came back: it re-reads the
file and compares it byte for byte. A git-based restore is checked with the tree
instead of the exit status of the command that was supposed to perform it, so
files that exist in neither revision now stop the analysis rather than being
dropped. Both failures are returned instead of noted, the differential's result
is recorded as an explicit analysis failure, and the report says the figures are
absent rather than zero.
The I/O entry points are variables so tests can inject a failure instead of
racing one: deleting either check makes TestRunMutationFailsWhenTheRestoreFails
and TestDifferentialPerTestFailsWhenRestoreFails reproduce the old behaviour
("a failed restore still produced a verdict: caught=1 total=1 survivors=[]").
key and hold_key need only the modal their keys go to, so they no longer
read the focused element's value. Typing is confirmed only when the
read-back differs from the value before by exactly the typed text put in at
one place, replacing what was selected there; the changed span is taken
from either end, since repeated characters make it ambiguous. A value that
merely changed and contains a short typed text, as autocomplete or another
writer can leave it, is unverifiable.
A modal's title and the name of the window it holds come from the
application. They were quoted but not bounded, so a 100000-character title
reached the model as about 100 KB, while every other label in the tree is
clipped to 160 by the helper. Both names now go through
textutil.BoundLiteral at that length: controls, bidi overrides and
invisible characters show as visible escapes, and so does a quote, so a
name cannot close its own quotation.
The Windows helper sent keys to whatever held the application's focus and
answered success. With Notepad's error dialog up, `type "hello"` returned
Done:1 while the text went to the dialog's OK button (#11366).
The helper now reads two structural facts, never a title or a message:
- a Win32 modal: an enabled window of the process whose owner chain holds a
window of the same process it has disabled (MessageBox, DialogBox, common
dialogs);
- a dialog drawn inside its window that declares WindowPattern.IsModal, which
is what Windows 11 Notepad shows: one top-level window, still enabled, with
a XAML popup holding its input.
With them:
- type and paste are refused as computer.blocked, before anything is sent,
when focus sits in a modal on an element that takes no text: a letter there
can be the dialog's mnemonic. Typing into a field of a modal (a file name
in a save dialog) still goes in;
- a ref or point step on an element in a window the application has disabled
is refused as computer.blocked, naming the modal that holds it;
- key and hold_key still go (Enter and Escape answer a dialog) and say which
modal they went to;
- type, paste and set_value read the focused element's value before and
after: confirmed when what was typed is in, suspected_noop when nothing
changed, unverifiable when it cannot be read (no value, a password, or a
value too long to compare). Values stay in the helper; set_value no longer
sends its read-back across the protocol;
- a snapshot lists the modals, and marks the modal's line.
The fixture gains a message box mode and an owned form with a text field;
the live tests run in the logged-in desktop session.
A computer_act step that the helper accepted read to the model as done,
whether or not anything reached the target. On Windows, typing into Notepad
while an error dialog held its input returned Done:1 and the text went
nowhere, and the dialog showed in the tree only as an unnamed window.
The helper protocol gains three optional fields, all read from structure:
- a step's result may carry an effect: a closed class (confirmed,
unverifiable, suspected_noop), the evidence kinds it rests on
(value_readback, value_unchanged), and the modal its input went to;
- a failure may carry blocked_by, with the new code computer.blocked for
input a modal would swallow;
- a snapshot may list modals: ref, title, and the window each one holds.
The kernel owns the grading contract: a class without the evidence that
proves it, an unknown class, or no effect at all (an older helper, or the
macOS helper until it reports one) settles as unverifiable. A step that
sends nothing, such as wait, carries no effect. computer_act says each
step's effect after its note and names the modal in a blocked failure; a
snapshot opens with the modals holding the application's input.
Tool schemas and descriptions are unchanged, so the cache-stable prefix is
byte-identical. Approval is untouched: grading only adds information, and the
new refusal only sends less.
The replacement now keeps the size and restores the mtime, so only the digest
comparison can tell it apart; it fails with the digest taken after the probe.
Refs #11434
The digest recorded for a persisted proof is taken before the launch, so an
executable swapped during the probe is not vouched for. The Probe field comment
now says the memo is private while proofs still persist.
Refs #11434
A failure with no sentinel on the read side (missing object, bad content
reference, size or digest mismatch) was reported as a failed copy. Reads now
default to corrupt, writes are classified separately, a folder without a
manifest is not listed as a session, and each copy button names its session.
Bash discovery launched each Git Bash candidate serially (10 s each), then a
PowerShell check, and remembered the proof only in process, so every cold start
repeated it in front of the Studio handshake. Candidates are now probed with a
hedge, the search shares one budget and falls back with probe_timeout, and a
successful launch is persisted keyed by path, size, mtime and content digest
(treated as a miss on any doubt). Boot gains shell, prompt and environment
phase marks so the start-up trace names every stretch.
Refs #11434
The free-text box of a question card stayed display:none until Other or a
single pick was chosen, so a card with only fixed options showed nothing
typeable while the run line read "Waiting on you" and the open plan kept
turning its in-progress arc.
The box is now always present on a live card and typing into it is an
Other answer, so there is one control for the free-text path. The run line
names the wait ("Waiting for your answer") and the plan stops turning its
arc while a decision is pending.
The legacy import dropped unreadable sessions behind event.Discard and
returned only a warning count. The migration layer now produces one skip
record per session (name, source path, reason code taken from the reader's
sentinel errors), the import response carries them, and the storage panel
lists them with the fact that the originals are untouched.
Problem: The browser contrast guard timed out opening Settings after the
duplicate sidebar footer was omitted beside visible icon navigation.
Root cause: The icon rail's Settings and Account controls did not carry the
chrome.settings and chrome.account identities declared for those actions.
Fix: Declare each footer action identity alongside its icon, label and
destination, and render it on the rail button. Extend the interaction test to
verify those identities and destinations through both footer locations.
Verification: The new identity assertion failed on the previous rail and
passed after the fix. All 36 related interaction tests, production and perf
builds, ui-census invariants and 10 fixture reports passed. The CI browser
guards (intrinsic, contrast and wheel) plus perf/side.mjs passed. Formatting,
go vet, make lint, make check, builtin/boot Go tests and diff checks passed.
A rule such as ask=["rm"] parses but its tool name matches no tool, so it never
fires and the user believes a boundary exists that does not. Saving a rule whose
tool no registered tool (or MCP/extension namespace) can match is now refused
with permissions.rule_unknown_tool; rules already in the file are kept, listed
as dormant in the permissions panel, and announced once at load with a typed
permission_rules_dormant notice.
Store.roots canonicalised every directory up to three times per call, and
BuiltinSubagentTools ran a full discovery pass for each of the five built-in
skills it checks. On Windows each canonicalisation is a chain of directory
lookups, so every boot.Build paid for around 120 of them in this one step.
Refs #12521
resolveTurnImages and withVisionRouting loaded the full project config on
every turn to decide how to route images, including turns with no image at
all, where neither answer is used.
Refs #12521
internal/assembly/boot took 478-717s on the Windows runner across 12 push runs, so the 12m per-package timeout panicked. The full step now tees go test -json and, on failure, prints the failed tests' output and the 15 slowest boot tests.
tool.after can report a finished write as failed, and a move can stop
half-done, so a failure is not proof that nothing was written. Non-prose
targets of a failed named-path call now keep the generic check owed. The
declared-checks cause also covers checks captured at task start.
Refs #12519
The crash files are the runtime's raw output and are not redacted, only
shell.log and host.log are. The notice now follows the same language the
page draws in instead of the machine locale.
A Studio kernel that died of a runtime fatal left no readable trace: its
stderr sat in host.log, which request logging rotated away every few
minutes, and the shell quit without recording how the kernel ended.
Routes that are polled declare themselves with routine(); a successful
request to one logs at Debug, a refused one stays at INFO. The kernel
mirrors runtime crash output into a bounded crash directory the shell
names, the shell quotes the tail into shell.log and the next launch tells
the person once.
Failed non-named-path mutations leave the extent of the change unestablished,
so the prose-only waiver no longer applies. The ask now names an unestablished
extent and declared checks as causes, spells paths from the workspace root, and
the trace fixtures are regenerated.
Refs #12519
The waiver read the whole workspace scan, so a README edit in a repository
with any code was still held to a verification command. It now reads the paths
the task changed: all prose (.md/.rst), scope established, no declared check.
The ask also names the changed paths that keep it owed.
Refs #12519
Git before 2.31 echoes the unknown option as an extra output line, which
Open reported as not-a-repository. Normalise the git dir, common dir and
work tree in Go instead, and do the same for the index path lookup.
The refusal echoed the command verbatim, so a long or multi-line command
could flood the message or forge a Cause line. The subject now goes through
textutil.BoundLiteral. A compound statement whose subcommand is dynamic
(git $X; ls) is attributed to shell expansion instead of leaking the
internal placeholder as an unknown program. The declared-writer cause no
longer claims a tool declared that it changes state.
Refs #12508
The install preview always shows every high-risk action, and an entry
with "tier": "eager" is high risk. A marketplace is capped at 64 plugins
and a skill repository at 200 skills, but a .mcp.json had no cap: a
package could ship tens of thousands of eager entries, bounded only by
the 2 MB remote read, and every one of them would be displayed.
parseMCPJSON, which planning runs for both a local and a fetched
.mcp.json, now refuses a file declaring more than maxMCPJSONServers (64,
the marketplace cap) with ErrInvalidManifest, before any entry is turned
into an action. The refusal is a hostFactError, so the caller reads the
declared count and the limit rather than an omitted detail.
With kind=mcp, a fetched URL whose path names a .mcp.json now returns
its parse error, as a local .mcp.json already did, instead of
ErrUnsupportedKind; the over-limit refusal otherwise never reached a
remote caller. Any other document fetched with kind=mcp, such as a
SKILL.md, still answers ErrUnsupportedKind.
Refs #12488
The planning gate recorded every call it could not prove read-only as a
side effect and told the model it "would change state", with no command,
no cause and no alternative. The effect is now Unclassified when the host
only failed to prove a read, SideEffect when a tool declares itself a
writer or an argument makes a reader write, and the refusal carries the
classifier's structural proof (unknown program, shell construct from the
parser, writing arguments, background, undeclared tool). The marker now
defines read-only shell in the same terms.
Refs #12508
A pack could declare recipes.density and recipes.corners, the schema
published both tiers and eight of the nine shipped packs carried them,
while nothing in the tree read either one. A pack asking for compact
density loaded clean, warned about nothing, and did nothing.
The reader keeps the two tiers now and reports a value outside them the
way it reports a token that fails its grammar, so an author sees the
problem in the picker rather than a pack that stays default in silence.
A recipe is a factor, not a size: the stylesheet owns what a radius or a
gutter measures, and a pack only says how far it steps. corners writes
--r-scale and density --d-scale, and each declaration that takes one keeps
its own value as the default, so a window with no pack on it is laid out
at exactly the numbers tokens.css writes.
Recipes is a hand-written wire mirror like the pack's other blocks, so it
joins the pairs declared in tools/repolint/wireparity.go: a field added on
one side alone then fails the gate at the line that cannot read it.
1.x suspended to the shell on Ctrl+Z, as terminal programs do; 2.x left
the key unbound, so it did nothing and there was no way to drop to the
shell without quitting.
Ctrl+Z now returns tea.Suspend from shortcutKey. Bubble Tea releases the
terminal (alternate screen, mouse tracking, bracketed paste) before
stopping the process and restores and repaints it on fg, so nothing of
1.x's manual mouse reset is needed. On Windows Bubble Tea has no job
control and ignores the request. GUIDE.md and its Chinese version list
the key.
Refs #12144
After scrolling back through the transcript, Enter on an empty composer
did nothing; getting back to the live output took Ctrl+End or the wheel.
1.x jumped to the bottom and followed the tail again on an empty Enter,
idle or while a turn ran.
An empty (or blank) Enter now does the same in full-screen mode, through
followTail, which Ctrl+End also uses. It still sends nothing. Inline mode
has no in-app view to move and is unchanged. GUIDE.md and its Chinese
version note the behaviour on the Enter row.
Refs #12144
Layout spacing carried 2364 distances across 48 distinct values, 1046 of them
on no scale at all, 10px 220 times and 6px 214 times. A pixel is not a level,
and those two are what a six-step scale could not hold: they are the values a
migration would have had to move 63% of the file's distances to reach. Eight
steps leave 44% off scale instead.
The tie rule is arithmetic rather than taste. 3, 5, 7, 9, 11, 14 and 20 each
sit exactly between two steps, and each sits above the step below it, so a tie
returns to that step and gives up only the air that was added by hand.
The scale is declared here; one surface draws from it. The activity group's
fold header takes --sp-6 for its inset and --sp-4 between the mark and its
label, and both were already those numbers: the two new steps renumbered the
ones below 12px, so the surface was renumbered with them and nothing moves.
TOKENS.md carries the scale, the step each role takes, and the tie table. Its
Space items were rewritten to fit the file's prose budget, which the first
draft of this change went one unit over.
A reply in Persian or Arabic was laid out left to right, and a sentence that
mixes in Latin words such as npm or a variable name came out jumbled, because
every block inherited the window's LTR base direction.
Markdown paragraphs, headings, lists, blockquotes and table cells now carry
dir="auto", so the browser takes each block's base direction from its own
first strong character: an RTL paragraph reads right to left and aligns to
the right beside English ones, and the bidi algorithm orders embedded Latin
runs correctly. Code blocks are left out and keep the document's direction.
The user's own message row and its edit box get the same treatment.
List indentation and the blockquote rule use logical properties
(padding-inline-start, border-inline-start), so they move to the right edge
for an RTL block and are unchanged for LTR.
Fixes#12114
Swapping in this build's system prompt after a model switch made the
transcript differ from the 1.x log, so the save that follows moved the
conversation to a new session. The system prompt is host contract, not
conversation; the unchanged check now leaves it out.
Refs #12474
The list scrolls inside a height the composer's aux area caps, so an editor
grown to its text could be taller than the list and get cut off at an edge.
The fit hook now takes the room left by the list, and the edited row is
scrolled into view. The browser check asserts the whole row sits inside the
list's visible area at desktop and phone width.
Refs #12477
The effort field is now resolved from the entry's reasoning protocol rather
than its kind, so an auto entry that resolves to a bespoke shape (Zhipu,
MiniMax, LongCat, DeepSeek) no longer claims reasoning_effort. The catalog
carries the protocols the field holds under and the panels read it instead of
keeping a copy. Adds the Anthropic none case and an end-to-end check that
existing effort-only configs keep their request bytes.
The harness measures context cost and could not report its own. It handed the
agent event.Discard, so the usage every billed completion already reports went
nowhere, and a batch's spend could only be read from a vendor console: this run's
index batch cost 604,544 prompt tokens (527,232 cached) plus 12,600 completion
tokens over 24 runs, and nothing in the tree would have said so.
The bench now installs a sink and totals what the agent reports — the same record
the product's own round path, compaction, command routing and failure paths bill
to. One path, not a second count of the same fact.
Two things stay honest about what the number means. A dry run prints an empty
suffix rather than "prompt-tok=0", because an unmeasured cost must not read as a
measured one. And the count is called usage_reported_requests: a request that
failed before reporting usage leaves no event, so it is not a count of attempts.
The package doc still said a cut of 50 actions sets previewTruncated, and the
field comment said escaped text sets it. The test-only PreparePlugin option
is marked as such.
The substrate the Status section asked for now has a result, and the result has a
direction, a size that is not yet quotable, and a reason.
The direction reproduces: six of six tasks, and the pooled paired batch searches
less and pages in less with the cue visible. A size is visible on the four tasks
whose absent-cue arm stayed out of a loop (44% less prompt cost), and it is
recorded as a post-hoc subset rather than an effect size: the two excluded tasks
are excluded because their arm looped here, and the qualification bar does not
screen for that. Saying 44% without that qualifier would mean picking the sample
after seeing the result, which is the mistake that invalidated the previous round.
The counts are stated separately because they differ: 36 runs executed, and the
token totals belong to the 24-run index batch that carried the usage line.
#12389 and #12035 each added the tsLookFile constant and an appearanceView mirror; the two merged cleanly as text and left tools/repolint unable to compile (tsLookFile redeclared). Keep one constant and one appearanceView entry, plus the zoomRangeView entry #12035 added.
PreviewTruncated also fired when only whole steps were left out, so the
confirmation said text had been shortened when none had. It now means that
text was cut or escaped, and HiddenActions carries the step count. The card
no longer repeats the plan-level sentence; it shows a short marker aligned
with the other rows.
The protocol table now declares where a chosen effort level lands in the
request body (reasoning_effort, reasoning.effort, output_config.effort). The
protocol catalog and the provider list carry it, and the add, detail and edit
panels word their help and example body from it instead of naming the Chat
field for every wire.
The Responses wire now takes the same thinking switch, reasoning_protocol and
declared levels as Chat in all three panels and in the edit endpoints.
reasoning_protocol=none is honoured when resolving the effective effort, and
a declared openai protocol lists its own ladder as the request vocabulary.
The shrunken-viewport case is renamed to say it checks the box stays within
60% of a shorter viewport with the caret at the end. The queue editor check
now compares against its computed cap and the room the list leaves. The
editor's own max-height was dead (the composer's textarea rule wins), so it
is removed again.
prefer=powershell tries Windows PowerShell 5.1 before pwsh, so advising it to a pwsh 7 user would trade away && chains. Advise prefer=pwsh when the interpreter in use chains.
Auto-detection never looked at where Git for Windows records its own install, so a Git outside the standard roots whose git is not on PATH was invisible and the session silently ran PowerShell. Add that registry location as the last bash candidate source, make a failed probe say whether it timed out, give the probe 10s instead of 3s for a cold start, and write the fallback warning that the startup notice already carries.
The theme category check had no test that reached ErrNotTheme. A marketplace
of fifty theme packages plus one with a hook must be refused and keep the
hook step visible in the envelope; sixty pure themes must pass. A table
sets each count field and the runtime alone, so dropping any term from the
sum is caught. install_source gains a test-only PreparePlugin option so the
market layer can plan a marketplace without a network.
The theme gate read the bounded preview, which lists at most 50 steps while
a marketplace may hold 64 plugins and a skill repository 200 skills, so a
hook or runtime past the cut skipped the check and the confirmation. The gate
now reads the unprojected actions; the preview keeps every high-risk step,
caps the rest, and reports how many it left out; the real skill count is
shown. Package inventory returns the package's own text again, and the
display projection is a separate call.
The FULL TRUST reason embedded the runtime command inside a sentence that was
then stripped as prose, so an unterminated escape sequence in an argument
swallowed the warning. The command is now rendered literally after the host
sentence. BoundProse reports removal as well as truncation, BoundLiteral
cannot be mimicked by typed escapes, and colliding display keys are flagged.
Stack on the refreshed read-only branch chip. Guard checkout atomically against all active panes and unfinished jobs in this host, preserve queued input cancellation, honor existing writer leases, and protect ignored local files including late-created collisions.
Problem: Usage, feedback, account/workspace and settings appeared twice when
the global icon navigation and workspace sidebar were both visible.
Root cause: Sidebar rendered its footer unconditionally, independent of the
navigation visibility already owned by App.
Fix: Pass the existing navigation visibility into Sidebar and render the
footer only when navigation is hidden. Preserve the fallback entries in off,
collapsed-only and narrow-window modes, plus existing feedback and settings
routing. Update layout/parity docs, pending release notes and fixture images.
Verification: 36 relevant frontend interaction tests, 11 release-note/link
tests and the frontend production build passed. Formatting, go vet, make lint,
make check, builtin/boot Go tests and diff whitespace checks passed. Browser
fixtures cover light/dark wide windows and narrow drawers.
The queued-line editor took its height from the number of line breaks, the
same defect as the rewrite box, so it now uses the shared fit-to-content hook.
Both height caps use dvh (with a vh fallback on the same line) so a soft
keyboard shrinking the viewport shrinks the cap. The browser check gains the
queue editor and a shrunken-viewport case; the harness can read a seeded
queue line back.
A failed job's notice now carries the same JobNotice payload as finished and
killed ones, plus the job function's own error text, so a frontend words
it from the label or id and shows the raw error beneath. A failure stored
without a payload still renders its text. Adds an emission test for the
killed notice.
Rebase the read-only branch chip onto the current Pane split. Keep the stylesheet within budget, register the complete workspace Git wire contract, and verify refresh after turn completion.
The rewrite box took its height from the number of line breaks, so a long
paragraph written as one line showed about 60px of it. The composer's
fit-to-content logic is now a shared hook used by both, with the box capped
at 60vh. A browser check covers a long single line at desktop and phone
width in both themes.
The completion notice carried only English text with a numeric id, so a
frontend could neither localize it nor name the job. Finished and killed
notices now carry a code and a JobNotice payload (kind, id, label); a
failure carries a code and keeps its diagnostic. Studio and the terminal
UI word them from the code; uncoded notices from older sessions still
render their stored text.
The folder mark beside the project name is decoration: it carries no click and
no meaning. A project can already be shown in the file manager, but showing it
reveals it in its parent folder; entering the folder itself needs a second
step.
This adds that step and wires it to the mark. The kernel answer and its refusal
are the ones reveal already asks for (/host/workspaces/locate), so the page
still names no location of its own and only the shell verb differs: openPath,
called on a folder the kernel itself named, instead of showItemInFolder. The
mark stays disabled when the shell cannot open a folder, or when no pane names
a workspace.
A text-less compaction_progress frame now tells a frontend the streamed
digest was discarded, so a retried summary does not append to the cut one.
The pressure effect test counts attempts as pairs with a larger retry cap
instead of capping raw request count. Adds a shared-window room test and
updates SPEC.md.
The summary request is capped at 16K output tokens and reasoning shares that
cap, so a model with no low reasoning tier can spend all of it and have the
digest discarded. Truncation is already a typed finish reason, so the host
retries once with the cap doubled (bounded by the model's configured output
ceiling) and fails with the cap and retry recorded when that is cut too.
The rail clipped itself to its animated width, which cut off the hints that
hang outside it, and shared a z-index with the workspace rail that paints
later. The rail now keeps its open width and slides off screen instead, and
sits one layer above the workspace rail.
Refs #12462
Problem: Default light Studio uses nearly identical frame and reading surfaces,
medium-weight prose, and a broad composer shadow, making long answers feel dense.
Root cause: Reading panes share the page token, replies inherit the small-UI
weight, and the composer uses the generic light surface and shadow recipe.
Fix: Separate the default light reading and composer surfaces, use regular
weight for default reply prose, and shorten the composer shadow. Preserve
explicit weight choices, installed theme packs, wallpaper surface overrides,
dark styling, and layout. Consolidate the duplicate light palette and document
the tokens. Include mock-fixture before/after screenshots at both viewport sizes.
Verification: 68 focused frontend tests; TypeScript/Vite build; browser theme,
weight, wallpaper, draft-input, and geometry checks; gofmt; go vet ./...;
make lint; make check; go test ./internal/tools/builtin/ ./internal/assembly/boot/.
Only the walk skips VCS store entries. An explicit write into a VCS store is a
change like any other. The snapshot-store exclusion is a path convention for
de-noising, so a symlink anywhere on the way disables it.
The unverified-change gap carried a sentence the kernel wrote in English, so
every language showed it; it now carries only the declined command. Changes the
host or a tool owns -- the task-snapshot store, VCS stores (including a .git
file), and directories tagged with CACHEDIR.TAG -- no longer count as work.
Refs #12457
The tests that fake Windows name probing on every OS spelled PATHEXT in
upper case while their fixture files are lower case. Windows file lookups
ignore case, so the real lookup finds mcp.cmd for .CMD; a Linux filesystem
does not, so under the fake the probed name missed the fixture and the
tests failed there. The seam fakes the probing, not the filesystem: spell
the extensions as the fixtures are.
Adds the off-Windows counterpart: where nothing probes extensions, a .cmd
beside the command does not affect the binding, and the command file does.
The shell exposed no way to leave the GPU out of it. Chromium fixes its GPU
use before the app is ready, so a saved choice (rx-hw-accel, in the existing
window-prefs.json) is read right after the profile directory is set and
before ready; the shell reports what this launch applied and what Chromium
ended with through a new shell:graphics handle. Appearance gets a Graphics
block, shown only in the desktop shell, saying which mode this launch uses
and that a change applies after a restart.
A fixed entry is upgraded from next to a concrete version only when a merged
closing pull request is contained in a studio tag whose release page is neither
a draft nor a prerelease. Prerelease status is read from the releases API, not
from a version cutoff.
A pack with a sky drew a blurred cloud bank under six backdrop-filter panels,
so every frame re-blurred 41% of the window and the idle window held a GPU
engine near full on integrated graphics. Frosted panels that are already
72-94% opaque blurred nothing visible, and the grain layer forced a
full-window overlay blend on every frame.
- Sky: CSS breathing animations removed; the canvas redraws on a 15 fps
timer and only while the window is shown and focused (presence.ts).
- backdrop-filter kept on the composer and the transcript backing only;
the other panels get a higher flat tint instead.
- Grain drops mix-blend-mode.
- New Appearance setting, Visual effects: Full / Reduced (rx-effects,
immediate, per machine). Reduced removes remaining blur, text glow, the
grain and the sky.
- perf/effects.mjs guards all of it in a real browser.
Clicking .html/.htm in the workbench explorer shows the file as a page in the
shared browser through the existing disk-path open, reloading when the file
changes on disk. Windows that draw no browser views and remote workspaces keep
the source editor.
The model menu now always opens with a search field, groups models under a
sticky provider heading with a letter chip, uses compact one-line rows and
grows with the window (up to 640px) instead of a short fixed box. Rows keep
the wire format and show a reads-images badge where the model declares it.
Enter no longer lands on the manage action, and an empty search says so.
An enable recorded for a server that .mcp.json or a project reasonix.toml
declares held for the server's name. The row now records a digest of the
declaration as written (type, command, args, env, url, headers), the project
.env values it expands, and the content of the workspace files it names:
its arguments, and every file the launcher could start for its command,
found by the launcher's own lookup (base/proc): the declaration's PATH for a
bare name (every Windows spelling of it, in key order), and PATHEXT
siblings on Windows, dotted names included. A different declaration reads
as ActivationChanged: off and awaiting the user. Rows written before this
carry no digest and ask once.
- Specs for a repository-declared server carry a LaunchCheck run before
every start and redial; env overrides merge in key order.
- Why a server waits is a typed code (awaiting_user_decision,
changed_since_enabled) on MCPHealth, the serve and ACP wires and the TUI;
Studio and the TUI word it in the reader's language; Studio shows the
launch line beside the switch.
- A connect or reconnect by name refuses a repository server in any pending
state with ErrMCPApprovalOwed (serve: mcp.approval_owed) and its launch
line; a server from the user's own files is still answered by a connect.
- Build reports each held project server once, so `reasonix run` prints
one warning line per server; serve prints the same at startup; /mcp and
`reasonix mcp list` show them. Repository-chosen names are sanitised
wherever they are printed, and run diagnostics sanitise every line.
- `reasonix mcp enable` prints the approved launch line; a project install
records the decision for the declaration as written; doctor --live
starts only what a session would.
Every test for the project MCP approval change, ahead of the change:
- boot effect tests through the real assembly and the settings switch: a
rewritten declaration, a launcher changed mid-session and an executable
swapped behind the declaration's own PATH must not start on an earlier
enable; held servers are reported once each at build; a repository's
server name reaches no notice as control sequences.
- config: the enable covers the declaration, the .env values it expands,
the files it names, the executable the launcher resolves (declared PATH,
every spelling of PATH, PATHEXT siblings including dotted names), legacy
rows ask again; base/proc holds the launcher's lookup.
- plugin: the launch check runs before start and redial; env merges in a
fixed order.
- control/cli/tui/frontend: a connect by name never approves a repository
server that waits; why it waits is a typed code on every surface; /mcp,
`reasonix mcp list`, serve's startup warning, run diagnostics and the TUI
panel print a hostile name as plain text; doctor --live starts only what
a session would.
The form moved focus from the tab into the body in a frame callback queued from the environment promise. The fieldset is disabled until that environment is committed, so a frame that fired before the commit called focus() on a disabled textarea and focus stayed on the tab. The move now runs in an effect on the environment, after the commit.
The unread interaction tests emitted wire events through a listener map that was only filled once the pane subscribed, and a missing listener was a silent no-op. The first test of the turn-ending group waited for the port to exist, which is earlier than the subscription, so an event could be dropped under load. emit now waits for the listener.
docs/GUIDE.md lists Ctrl+Enter (or /steer) as the key that adds guidance
to the active turn, and 1.x bound it so. 2.x bound only Ctrl+S, so on a
terminal that reports Ctrl+Enter the documented key did nothing.
While a turn runs, Ctrl+Enter now sends the composer as steer through the
same path as Ctrl+S. Idle it does nothing, as in 1.x, so a press meant as
a newline never submits the draft. Ctrl+S is unchanged.
Refs #12144
With in-app mouse capture on, a right-click copied an active transcript
selection and otherwise did nothing. 1.x pasted the clipboard's text on
a right-click with no selection, and docs/GUIDE.md and docs/CLI.md say
2.x does too, so the documented paste never happened.
A right-click with no selection now pastes through the same path as
Shift+Insert: text only, through the normal paste handling, with the
existing hint over SSH. It pastes only while the composer is drawn; the
condition that draws it moves into composerShown so the view and the
mouse read one answer. A right-click on a selection still copies it.
Refs #12144
Up and Down recalled history only while the composer held a single line.
Once a recalled entry had several lines, Up moved the caret to its first
line and then stopped: nothing older could be reached, and a multi-line
draft could not step into the history at all.
Up now recalls from the composer's first line and Down from its last, as
1.x did; between them the keys move the caret as before. A single line,
soft-wrapped or not, behaves exactly as it did.
Refs #12144
Cause: 30 days of Studio UI changes landed without screenshots or review, and a visible control was hidden without a linked issue or a release note.
The PR template had no UI requirement and no table tracked 1.x features.
Refs #12427. The rail moves left of the sidebar, gains the feedback entry, and the Appearance setting becomes always / only when the sidebar is collapsed / never. Stored on and off values keep their meaning.
The /mcp panel opened a server's tools with Enter and backed out with
Esc only. 1.x's MCP manager also went in with l / Right, backed out one
level with h / Left, and closed from anywhere with q, and docs/GUIDE.md
(both languages) still lists those keys, so pressing them did nothing.
The panel now takes l and Right as Enter, h and Left as Esc (back to the
list, or closed from the list), and q closes it from the list or a
server's tools. The repository-declared confirm line is unchanged: any
key but y still cancels it.
Refs #12144
The ask card took only the arrows: Up/Down (and Tab) to move between
rows, Left/Right to move between questions. 1.x's question card also
took j/k and h/l, and docs/GUIDE.md (both languages) lists them for it,
so pressing them on 2.x did nothing.
The card now takes k as Up, j as Down, h as Left and l as Right, with the
existing wrap, clamp and auto-submit stop. A typed answer or note is
unaffected: the composer gets those keys as text before the card does.
Refs #12144
The approval card moved its cursor with the arrows and Tab only. 1.x also
took j/k and Ctrl+N/Ctrl+P there, and docs/CLI.md says approval rows use
the same row-selection keys as the pickers, so pressing them did nothing.
The card now takes k and Ctrl+P as Up and j and Ctrl+N as Down. None of
them is a row shortcut (y/a/p/n/x), so no answer changes meaning.
Refs #12144
A call reached through use_capability only learned its delegation profile after the target finished, so the deck's sub-agent list and the running count missed every sub-agent until it ended. The profile is now announced when execution starts.
A notes response is accepted only when its declared type is markdown or plain
text, so an error page served with 200 is refused and never written. The cache
name is lower-cased with a digest for mixed-case prereleases and over-long
versions are refused. Expanding a row on a narrow window keeps its label and
date. Error wrapping and the wait group follow the linter.
The kernel serves GET /studio/versions/{version}/notes from the release mirror,
keeps each fetched document on disk for good, shares one request between
concurrent readers and refuses by typed code. A row whose catalog entry names
notes gets an inline view with images off.
The notes are optional, so a failed upload or head-object warns and writes the
catalog entry without a pointer instead of failing the Mirror to R2 step.
Studio release workflow uploads studio/notes/<version>.md ahead of the catalog
write and gives the new entry an optional notes field. A one-time script
backfills the entries already in the catalog.
pendingAsk returned an arbitrary entry of the ask map, including one still queued behind the prompt lock. When the second ask took the lock first, the test answered the queued one and then waited for it while the other held the lock for the user: a hang until the package timeout.
useWallet treated the empty account of a pane whose status had not arrived as a source, so the first status turned it into a source change: the wallet read that had already landed was emptied and read again.
Update capped the lock wait at 5s even for a caller that supplied its own deadline, so the number of writers a queue could hold was fixed by the package rather than by the caller. The serve tests now pass a budgeted context and fail on the returned error instead of reading a swallowed one as a wrong count.
The host appends a background-job-finished notice as a user-role message once the job ends. When it landed before the CHECK turn composed, the provider read the notice as the last user message, took its default branch and never issued wait/read, so the answers it recorded for w1 and r1 were empty.
Every config load re-parses each enabled plugin package, and a settings sheet
fires a dozen such loads at once, each queued behind the package store's lock.
With a 293-skill package a burst costs one full parse per request. Callers now
wait for a parse that starts after they arrived and share it, so a burst costs
at most two.
Every type the desktop re-declares by hand is held to the kernel's, and
src/port/look.ts was the one file under port/ that no pair named. The theme
a person is looking at is all of it: the pack listing, the pack's two shape
blocks, the import receipt, and the reader's size and wallpaper.
Declared and not inferred: no spelling tells a mirrored contract from a
struct that merely carries json tags.
The desktop-sessions-v5 by-id store carries no workspace, so every session
from it landed in the workspace picked in the dialog. The owner is recorded
in desktop/workspace-state-v1.json; read it from the picked folder and route
each session to its workspace while that folder still exists.
Pane.tsx sat at the 800-line ceiling with twelve open pull requests editing
it. The view (flow, analysis, browser) lived in component state, out of reach
of the tab strip and of any setting that will place tabs.
The view now lives in state/paneview.ts, keyed by pane id and owned by the
mounted instance, so a remount for another conversation still starts on the
conversation. Submit, surface slots, the meter row, the plan fold and the run
line move to files of their own; props and the report type move to
panetypes.ts. No visual or behavioural change.
Adds css-important, css-size and stray-file rules to repolint, deletes the
stray Settings.tsx-e, runs the intrinsic, contrast and concurrent browser
guards in the studio workflow, and adds a report-only wheel-scroll probe.
The rail was hidden by the studio layer's prototype alignment, which left Nav.tsx rendered but display:none. It is drawn again, beside the workspace column, when this machine has it on (default) and the window is wider than the scene fold.
The BrowserWindow carried no backgroundColor, so any pixel the page does not
cover (frameless window at a fractional display scale) showed the default
white. The ground now follows the stored rx-theme, or the system for auto, at
creation and when preferences are saved or the system theme changes.
The sidebar folds a conversation's recovery copies under one row, but
/tree/sessions/remove deleted only that row's file. The next copy then took
its place, so the delete looked like a no-op with the order shuffled.
Remove the lineage the way archive already addresses it, leaving siblings a
pane has open.
A 400/422 on a request whose history carried reasoning the connection
could not replay was labelled "the endpoint wants the thinking content
sent back", whatever the endpoint actually said. Dropped reasoning is a
fact about the request, not a cause the host observed, and the endpoint's
refusal carries no typed signal separating the two. The hint is removed;
the refusal now shows the status guidance and the endpoint's own words.
Search and list dropped results silently: hits below the relative score
floor (KeepTopRelativeScore compacts in place, taking the count with it)
and everything past limit. A reader could not tell a complete result from
a truncated one, while the empty-result message already warns that no
result is not proof of absence.
Both paths now carry the note automatic recall already uses for its own
omissions, counting both reasons and staying silent when nothing was
dropped. The spelling lives in one helper the recall block calls too, so
the two cannot drift apart; its output is unchanged. The count is one
function as well, so the case where both reasons fire is tested directly
instead of only through fixtures. Empty results keep their guidance: the
relative floor always keeps the top hit, so empty still means nothing
matched.
Refs #11853.
File tools resolved a path through the filesystem before comparing it with the
workspace, so a path spelled as a network share was looked up first. On Windows
that lookup is itself a connection to the machine the path names.
A network path is now answered by its spelling alone, with no filesystem call:
it passes only below a workspace root that is itself a network path, compared
lexically and case-insensitively, and is otherwise refused with
workspace.network_path_outside_scope. The check is one function in fileutil and
sits at the entry of the read tools, the writers, their previews and write-path
resolution, write claims and path observation. Local paths and network
workspaces behave as before.
A path spelled as a network or device path is now answered by its spelling
alone, before any filesystem call, and is refused with its own code
(browser.network_path) instead of the generic outside-workspace one. The
check lives in fileutil next to the containment helpers and covers page
loads and file uploads.
Both boxes submitted on Enter regardless of pointer type; they now ask
touchKeyboard() like the composer does. The queue editor also ignores Enter
while an IME composition is open, as the rewrite box already did.
The anonymous ping was a single attempt at launch with a 1 s total budget; a
failed attempt dropped the day's claim and nothing tried again until the next
launch, and a host left open past UTC midnight was never counted again. The
reporter now retries a failed ping with bounded backoff, waits for the next UTC
midnight (plus a splay) after a success, and stops with its context. Consent is
re-read before every attempt, so turning the setting off stops the next send
without a restart. A refusal the server marks permanent (400/401/403/404/410)
is not retried for the rest of the UTC day; 429, 5xx and network errors back
off. Request timeouts are relaxed since the work is off the launch path. The
payload is unchanged.
The interactive session that carried the first-run consent and the reporter was
removed, and the replacement terminal UI never started either, so CLI 2.x users
were neither asked nor counted. runTUI now runs the documented consent on the
plain terminal before the UI takes the screen, starts the reporter only after a
mode is stored, and wraps the controller sink so the existing bounded counters
flow.
An ask or tool-approval wait kept the session's claim for as long as the user
took to answer, so other sessions were blocked on paths nobody was writing.
The claim is now yielded for the wait and taken again, queueing behind any
session that claimed the extent meanwhile, when the user has answered.
Refs #11531
The address bar read "D:/DevCode/x.html" as host "d" over http. A drive
letter plus a separator, a leading backslash pair or a single leading slash
now names a file: URL, parsed once in Go (browser.LocalPathURL, used by the
agent's checkURL) and once in the window (localpath.js), both held to one
shared table. The window opens local files for the person; a network share
is refused with its own code instead of a DNS error, and a page's own links
still never reach a file.
The summary request is one user message that ends on the last tool output, so
the model sometimes carries on the agent's work (pseudo tool calls, "next I
will ...") and the host installed that reply as the digest: a read-only fold has
no changed file for the coverage guard to compare.
- The user turn now ends with an explicit ask for the briefing.
- A reply with no level-2 heading, the shape the summary prompt requires, fails
with the typed code summary_not_digest. It is billed to the compaction, is
retried like any transient summary failure once the input has grown, and
degrades to the mechanical digest when the fold must free space.
- The code has frontend text in the TUI catalogues and the desktop table.
- Test doubles that answer summary requests now speak the heading contract.
The worker now states a profile on GET /v1/feedback/mine and names the
window in every feedback refusal. The kernel client dropped both: the
profile never reached a frontend and a 429 collapsed to "wait N seconds".
- feedback.Mine carries a validated Profile (null when absent or
incoherent; the last confirmed one while offline, cleared with the
install identity).
- Refusals decode error.params into a typed LimitError (window, reset
time, wait) for rate_limited, reply_limit and busy; serve forwards the
same fields through refuse.
- Studio shows the level row, progress track, limits, trust note and the
badge artwork (single-colour SVG on text tokens) in My feedback, and
says which window a refusal ran out of and when it resets.
- The terminal's /feedback list prints the same standing and refusals,
worded through the i18n catalogue.
Refs #12025
keepToolCallGroup kept every result of a parallel tool-call group when any one
failed. Models batch 4-11 calls per turn, so one failing test pinned the whole
batch of reads, the candidate exceeded the 50% checkpoint ceiling, the billed
summary was discarded, and the non-transient block was never retried.
- The retained assistant turn now issues only the calls whose results stay; the
folded copy carries the rest, so the summarizer and the coverage guard still
see them and no request issues a call nothing answers. Turns whose call ids
cannot be told apart stay whole.
- When the retained failures alone would exceed the ceiling, the fold gives up
retaining them before the summary is requested; the coverage guard still makes
the digest name each failed command.
- A candidate_above_ceiling block is released by input growth below the physical
ceiling, and an overflow on a changed input may fold, instead of holding
forever.
The Settings subagent row writes subagent_model, while an entry the CLI
writes under subagent_models is read first for its profile, so the row
looked inert. The kernel now lists those entries per role with the file
that holds them, the row names each one, and a user-config entry can be
cleared from the row.
Refs #12151
A brokered pane takes its default from the home machine's catalogue, so a
default_model written next to the pane's own kernel was never read and the
row snapped back. The default is now its own request, POST /default-model,
sent to this machine's pane; the pane's switch stays a plain /model. A
brokered pane refuses the request by code instead of ignoring it.
Refs #12043
agent.subagent_effort was validated only against the parent model and then
applied to whichever model a subagent selected (subagent_models.task, a
skill's own model, a per-profile model, a per-call model) through
NormalizeEffort, which remaps across providers or lets an unsupported
level through. The default is now resolved per executing model by one
function; a level that model does not declare is dropped with a warning and
the model keeps its own default. Explicit efforts still go through
NormalizeEffort, whose refusal now satisfies errors.Is(err,
config.ErrEffortUnsupported) and names the model and level.
The ACP-specific subagent resolver and task defaults had no caller (ACP
sessions are built by boot.Build), so they are removed.
Refs #11871, Refs #11872
The notice appended to bash output when a command mentions the session store
read like a refusal: it said to stop retrying and report, and when the command
matched nothing it was the only text returned. A model that had just been told
to read a spilled output gave the output up. State that the command was not
blocked and name read_file/grep for spilled outputs.
Cancel had no bounded wait, so a stop the kernel never acknowledged left the
button on "stopping" forever, and stopping only cleared when the running flag
dropped. Cancel now shares the bounded acknowledged post with decisions, which
fails with a typed DeliveryError (kernel_busy, unreachable, ui_stalled). A
deadline that fires far past its time on the monotonic clock is reported as
this window being stalled instead of a busy kernel. A stop also ends when the
kernel's status reports no live turn.
Refs #12232
The shell held nothing against system suspension, so a laptop that went to
sleep overnight froze every running session. The kernel now answers how many
panes have a turn in flight (GET /tray/running, local and remote, read from each
controller), and the Electron shell holds a prevent-app-suspension block for as
long as that is non-zero, reconciling against the answer on every poll so a
closed pane or a dead kernel cannot leave it held. A machine-level switch in
Settings > Appearance > Window turns it off.
Build windows/arm64 on a native windows-11-arm runner and carry it through the
signing chain as a second leg of every Windows stage: one tree per job, legs
serial under the certum-signing group, a failed leg reruns alone. The package
check learns app-arm64.7z and refuses any PE image that is not the leg's
architecture, a new install smoke runs each installer on its own architecture
before publish, and publish counts, names and hash-checks both architectures.
The arm64 installer is packed with the 7-Zip BCJ filter: the default ARM64
filter yields an archive the NSIS extractor cannot decode, and the installer
then exits 0 with every executable missing. Signing contract tests pin the
matrix, the serial legs and the secret-free stages.
Add a windows-11-arm leg to the Studio package job: the kernel and the
computer-use helper build for the runner's architecture, every PE image in the
bundle must carry the package's machine type (the installer-written
elevate.exe and uninstaller are 32-bit by construction), and the package is
started, installed and uninstalled where it is meant to run. An arm64
installer now refuses a machine that is not ARM64 Windows instead of
installing nothing, and a second job checks that refusal on an x64 runner.
The arm64 installer is built with the 7-Zip BCJ filter: 7-Zip applies an
ARM64 filter to ARM64 executables by default, the NSIS extractor cannot decode
it, and the installer then exits 0 with every executable missing.
The ARM64 leg runs on pull requests only when files it verifies change.
Build and verification only: no secrets, nothing is uploaded to a release.
With no browser tab, following a link asked the kernel for the link's tab
and the column, mounting at the same moment, asked for an about:blank of its
own. Which of the two the selection rules landed on depended on reply and
poll order, so the blank page was sometimes the one shown.
openPage remembers what is being opened until the kernel answers, and the
empty column takes that page instead of requesting another.
Refs #12251
The sub-agent chip summed every delegate in the session, finished ones
included, while the background-jobs chip counted only running jobs (and
fell back to the total when none ran). Both chips now show what is
running; the sub-agent popover reports "N running / M total" and keeps
the finished ones listed.
An optional extension whose sidecar is not running was reported as an
English sentence with no code, which the composer's notice bar clipped
at its edge and which named no next step.
dispatch now exposes ErrNoLiveSidecar and a SidecarDown hook; boot turns
it into a notice coded extension_skipped whose Detail is a typed payload
(extension, point, reason). The desktop words it from the payload with a
button to Tools and integrations and lets the bar wrap, the TUI words it
from the i18n catalog, and ACP chunks carry the code and payload.
Refs #12258
The form card's submit awaits onAnswer and clears its in-flight flag in a
finally block. The tests resolved that answer on a 20ms timer they never
waited for, so the setState could run after the jsdom environment was torn
down and vitest exited non-zero on 'window is not defined'. The tests now
wait for the answer to settle, and an afterEach fails any test that leaves
one in flight.
An entry parked as uncertain with code steer_unapplied (guidance the turn
never read, e.g. after a stop) kept the queue paused, and the panel offered
Retry only for blocked entries, so Resume was a no-op and the only way out
was withdrawing and resubmitting. Other uncertain codes may already have
run and stay take-back-only.
Rename the model-section action to "Read available models from provider" and
offer the same action beside the add-model field, with the failure shown there.
Light --warn is 4.58:1 on white and 3.86:1 on the floating layer, under the
4.5 text line. Text rules that set color to a bare --warn, --ok or --err now
use the matching ink token; borders, washes and glyph marks keep the base.
A structural test fails a new bare hue used as text.
useWallet now takes the account (connection) the wallet belongs to: a change
of account empties the wallet at once and reads again, and only the latest
read may land, so a quick double switch cannot leave the first account's
number up. Switching models within one account keeps the wallet and does not
re-read. An account with no wallet (204) or an unreadable one shows nothing
instead of the previous figure. A reading the source reports as unavailable
is drawn in the warn-ink token and carries a text label; the figure itself is
untouched.
Refs #12266
The updater and studio-manifest already key packages by GOOS-GOARCH; nothing
pinned what happens around a second Windows architecture. Cover both
installers resolving to their own keys, a manifest without the client's key
resolving nothing, a delta for another architecture meaning no delta here,
and document the platform-key rules in the release runbook.
TestEffectVersionControlRidesTheTurnNotThePrefix builds two controllers and
requires byte-identical system messages, but the environment section in that
prefix comes from the process-wide probe cache, keyed on PATH and shared with
every earlier test. The boot package runs longer than the cache's five-minute
TTL on Windows CI, so an answer an earlier test left can expire between the
two arms: the first arm reads it, the second re-probes, and a tool that timed
out once (a cold docker --version) appears in only one prefix.
The test now appends an empty directory no other test uses to PATH, which
keys a fresh probe answer for the first arm that the second reuses seconds
later. PATH lookup order and what each probe resolves to are unchanged.
A second TUI attached to a session kept an ask or approval open after the
turn ended without a decision receipt (the other screen skipped the ask or
cancelled the turn), blocking its input on a prompt the kernel had already
dropped. The desktop already seals these in sealTurn; the TUI transcript now
does the same on turn_done.
Xiaomi lists mimo-v2.6-pro, mimo-v2.6-flash and mimo-v2.6-pro-ultraspeed, and takes mimo-v2.5-pro and mimo-v2.5 offline at 2026-10-21 10:00 Beijing time with no replacement model.
A PreToolUse/UserPromptSubmit hook that could not be evaluated used to be
treated as absent: a malformed matcher never fired, a spawn error was
DecisionError (warn-level), and a payload json.Marshal error was dropped so
the hook ran against empty stdin and passed.
On gating events (and Claude-imported PermissionRequest) each of the three
now blocks, with a typed cause (ErrInvalidMatcher, ErrSpawnFailed,
ErrPayloadUnserializable) carried on Outcome and Notice. The model gets only
structured facts (code, step, event, scope, source, label), never the
script's output or the underlying error prose. Observing events still do not
block; they report the outcome to the user. Save rejects an invalid matcher
on tool-matcher events.
The tests that pinned the old behaviour (malformed regex "should not fire",
spawn-error => DecisionError on PreToolUse) are rewritten to the new contract.
101f2ca12 stopped composing the fact index into the cached prefix, but
SESSION_MEMORY_RETRIEVAL.md and REASONIX.md still said it joins the prefix
at session start. Only pinned bodies and standing instructions do; the
index is reached through the memory tool.
A cache hit never wrote the snapshot of the directory it was served to, so
when the 5 minute in-memory entry expired the next build probed live and
could render a different section, moving the cached prompt prefix. Adopt the
served answer into that directory's snapshot.
Also print the first differing bytes when the scheduled-run prefix test
fails, so a future flake is diagnosable.
The host appended a <context-budget> block to the turn tail at 75% and 92% of
the compaction trigger and raised a matching warning card. The model can still
ask: the read-only context_budget tool and its figures are unchanged, and its
description now says when to check, since nothing reports it unprompted.
The window opened a link in a new tab and then switched the active tab back
to the agent's. The page list was re-read between the two, so the panel saw
the agent's active tab move to the new tab and back, and its follow-the-agent
rule selected the agent's tab again: the link's page appeared but was not the
one shown. Session.VisitBeside opens the tab without activating it unless the
session has no active tab, so there is no transient to misread.
A prompts/get failure surfaced as the bare transport error, and the model
never learned the user had invoked a prompt. MCPPrompt now returns a typed
MCPPromptError (prompt, server, stage; matches ErrMCPPromptFetch) and owes
the model a turn-tail block describing the failed invocation.
Draw the kernel's typed `unread` on session rows (dot, bold title, spoken
name), as a count pill on folded project rows and as a dot on pane tabs.
A session is marked viewed through the pane's own endpoint when it
becomes the pane in front, when an unread row or pane is clicked, and
when a turn ends in the pane in front while the window has focus. The
mark is cleared optimistically and restored if the call is refused.
Refs #12230
/attachments refused with plain text, so the composer could only show
"Could not add" and left the reason in a tooltip. Refusals now carry
attachment.unsupported_image, too_large, empty, unreadable and write_failed
with params, the port keeps the code, the catalogues say each in both
languages, and the failed chip shows the reason with Retry.
Panes that are not on screen kept re-rendering on every streamed delta, and
each of those renders also rebuilt the reply actions handed to every settled
row. With nine sessions streaming the renderer's main thread saturated and
queued work (status reads, stop, settings) starved.
A hidden pane now takes the first token of each stream at once and coalesces
the rest into one batch per second; any other event flushes first and bringing
the pane forward flushes immediately. Reveal pacing, the thinking clock and the
wait clock only run for a pane that is shown. The reply actions keep their
identity while a streamed answer grows, and the folded thought's length is no
longer recounted per delta.
Refs #12232
MarkSessionViewed pre-checked the session sidecar with a plain unlocked
read. Concurrent views replace the sidecar, so on Windows that read can
meet it mid-replace (access denied, or torn JSON after the replace
fallback) and the view failed with a 500 instead of waiting it out. The
pre-check now uses the tear-tolerant read the other sidecar readers use.
Refs #12230
The kernel now records two optional facts in the session sidecar,
finished_at (stamped when a turn ends) and viewed_at (stamped when the
person opens the session). A session is unread iff finished_at is after
viewed_at; the sidebar tree and GET /sessions carry it as a typed
`unread` boolean. A turn the person cancelled is stamped as viewed.
Controller.MarkSessionViewed (POST /sessions/viewed) writes only the
session sidecar and takes no workspace or transcript lease. Sidecars
without the fields read as seen.
Refs #12230
A fold leaves the model with a digest, and the host re-projected only the
step ids of plan-seeded unfinished lists. A task list the model wrote itself
(no ids), a list with every item complete, the files changed and the checks
that passed all survived only as digest prose, which a weaker model reads as
work not yet done and repeats.
While a fold is installed the request now ends with one derived "Host
progress record": the canonical task list with statuses when no todo_write is
left in view, the files changed in the folded prefix, and the verification
commands whose latest outcome was a pass (marked when a file changed after
them). It is derived from tool receipts, bounded (30 files, 15 checks, with
the omitted count), memoised on the covered-prefix fingerprint, and ends with
the fold. Sessions that never folded carry nothing extra.
A live serve with no record is attached and recorded, launch no longer
deletes the port/pid files another serve published, and a live serve
whose provider source differs fails the connect with a coded error
instead of being terminated. Document what desktop exit does to the
remote runtime.
The lock directory sat at a shared, predictable /tmp path, so another
identity could create it first and every later run by this user failed
on os.Chmod with a bare EPERM.
Put it under the OS user's home (from the account database, so HOME and
TMPDIR overrides do not move it), as Windows already did. Keep the owner
uid check before the chmod as a backstop: it returns ErrLockDirWrongOwner,
naming the path and uid, without touching the directory. The /tmp name
is only a fallback when no home is recorded. The check is a no-op on
Windows; lockdir_owner_other.go is renamed lockdir_owner_windows.go.
Fixes#11372
Package segments made only of dots, leading dots, a leading underscore on
unscoped names and .tgz/.tar/.tar.gz names are no longer accepted by the
installsource and registry checks. parseGitHubRepoSource refuses dot-leading
owners and dot-only repos such as the one TrimSuffix(".git") leaves from
"...git".
Refs #11353
Co-Authored-By: Claude Sonnet 5.5 <[email protected]>
perf/tooltip.mjs looked the attach button up by its aria-label in each
language, and the English label it searched for, "Add attachment", is not
the shipped one ("Add an attachment"). The lookup found nothing and the loop
skipped on, so the four English attach combinations were never measured and
the guard stayed green.
Controls are now found by structure, data-action for refine and the attach
class for attach, and a control that cannot be found fails the guard rather
than being skipped. All 24 combinations (light/dark x zh/en x two widths x
three cases) are measured: 96 checks, where 80 ran before.
Refs #12156
The veil blurred the whole window behind it, which the compositor redoes every frame while anything beneath animates. A flat, more opaque scrim dims the same way for free.
GUIDE.md lists "Ctrl+L or /cls" as clearing only the visible transcript, and
1.x bound the key, but the 2.x terminal UI left Ctrl+L unbound. It now runs
the same clearDisplay as /cls: the conversation, session and the kernel's
task list stay, and a running turn is refused the same way. Being a key
rather than a typed command, it leaves a half-typed draft in the composer.
Refs #12144
The fold trigger compared only a local estimate against the threshold. That
estimate is characters times one tokens-per-character ratio, and the ratio is
not learned from requests that carry images or provider-side tool results, so a
session with screenshots can sit well under the real size and defer the fold.
The trigger now takes the larger of the local estimate and the provider's
reported prompt_tokens (cached tokens included) for the last request plus the
local estimate of what the visible view gained since. The anchor is dropped when
a fold changes the projection, when the transcript is swapped, and when the
usage is estimated, aggregated over retries, or padded by provider-run tools.
The hard ceiling and the overflow paths still read the local estimate alone, so
a provider count that no fold can lower never ends a turn.
The slash and @ completion menu and the /resume picker moved only with the
arrow keys, while 1.x also took Ctrl+N / Ctrl+P there and GUIDE.md and
CLI.md list both for these lists. The provider and quick pickers already
bind them. Both lists now take Ctrl+N as Down and Ctrl+P as Up; neither is
text, so a picker's filter is untouched.
Refs #12144
bash_output and wait reported a background job that later failed or timed
out only in their text, with no tool error. The activity panel and the
failure count read the host-set error, so a job that failed after it started
never turned red and was not counted, while the model was told it failed.
Both tools now return errJobFailed, naming the failed jobs, whenever a job
they report ended as failed (a background command's non-zero exit and its
timeout both end that way). The output is returned alongside, so the model
still reads it under the error line. A job that finished, was killed, was
interrupted or is still running is not a failure and the call stays clean.
Fixes#12201
POST /providers/probe built its refusal message from probe.Error(), which
carries the model-list response body verbatim. That body can echo the
request, and this route receives the key typed into the add-provider draft,
so a 401 that quoted the key sent it back in the refusal. A failure that was
not a ProbeError came back as an untyped 502 with the error's text.
The message now carries the probe reason and only what endpointDetail lets
through (one bounded printable line, the draft key and credential-shaped
text redacted), the same treatment /providers/check gives its detail. A
failure without a probe identity is a kernel fault and refuses as 500
provider.probe.failed without its text.
Fixes#12188
A phone using the web client got an empty Workbench when the agent opened a
page: useBrowserTabs returned early because the web host draws no native
views, so the tab list was never read. Read it on every host and, where the
page cannot be drawn, show its title and address as plain text with no
controls.
Opening a session pane read GET /checkpoints three times against a real
kernel: the mount effect read it, the turn-boundary effect read it again on
its own first run, and a third time when the first status named the session
path.
useCheckpointRefresh now owns every read outside a session reload: the mount
read, a move from one known session to another, and both turn edges. The
kernel opens a turn's checkpoint when the turn starts and fills it while the
turn runs, so the read at turn start stays and the rewind entry still appears
on a running turn's message. The first status naming the session the mount
read covered no longer reads again, so opening a pane costs one request.
A failed read is tried once more after a second, since no later edge is owed
to it, and a read that a newer one superseded is dropped instead of
overwriting the newer list.
Refs #12080
The save area sat at the end of a long form, so its state was out of sight
exactly when it mattered. The bar is now sticky, shared by the edit and add
forms, and holds the failure or not-applied reason beside a single persistent
status line with a dot for each phase (clean, unsaved, saving, saved, saved but
not applied, save refused). Ctrl/Cmd+S saves a pending draft.
A skill that declares paths: is held out of the per-turn listing, the
capability catalog, the slash_command listing and the unknown-name hint until
the session has touched a file its globs match. The set of touched files is
owned by the skill store; eligibility is memoised against it and dropped when
it changes, and the next turn after a match owes the listing again. The user's
/<name> and run_skill by exact name are not gated.
The shell card and the activity group decided failure a second time from
execution.state and exitCode, so state background_started (no error, no exit
code) rendered a red BACKGROUND_STARTED badge and counted as one failure.
The host already classifies every call once: err is set for failed,
timed-out, cancelled and refused calls. Read only that.
Verifying a model was one click per row, and the form blocked while any one
ran. The model header gets a Test enabled models button that runs the existing
per-model check over the ticked rows, two at a time, and stops as soon as the
draft changes or the form is left. The summary line counts rows still being
verified, and the per-row live regions are dropped so a run is announced once
through the summary.
The agent tells a PathObserver which workspace files a completed call named,
through the ReadTargeter and WritePathResolver capabilities its tool already
declares; failed, refused and hook-blocked calls feed nothing and sub-agents
are not observed. Swapping the conversation rebuilds the set from the
transcript. skillSet wires its PathHits in; the catalog does not read it yet.
A failed catalogue refresh surfaced under the form's footer and a silent
refresh that found nothing looked like a click that missed. The add form's
connection failure also rendered below its save button.
Refresh failures and outcomes (including 'no new models') now render under the
model header, verified rows are summed up in one status line over the ticked
models, and the add form's connection failure sits under the button that ran
it. Results clear when the address or key they depend on changes.
Skills may declare file globs in a paths: key. The registry now parses them
(rejecting negation, parent traversal and malformed globs into InvalidPaths,
which doctor reports) and a PathHits set records workspace files the host saw
touched, resolving symlinks so nothing outside the workspace counts. The set
is owned by skillSet; the catalog does not read it yet.
Leaving the edit form for another service, the Add form, another service's
rename, or a protocol switch dropped pending edits without a word.
EditConn reports the pending state it already computes; Providers holds it
and routes every way out of the form through one guard that asks to keep
editing or discard. A save in flight is not interrupted and not asked about,
and the question closes by itself once nothing is pending.
The edit form had no way to tell pending edits from stored ones: Save was
always enabled, Cancel was a remount in disguise, and a successful save
silently turned into the same form with nothing said.
Pending state is the fingerprint of what a save would send, compared with the
one taken at mount, so typing a value back or toggling a row off and on is
clean again. Save is disabled when clean, Cancel becomes Revert, and the footer
says whether there are unsaved changes, no changes, or that the save landed
once the list has been re-read.
Cause: POST /providers/check returned a free-text `error` assembled from the
probe error's message. The probe already knows whether a key was refused, a
path is missing, nothing listed chat models or the deadline expired, but that
identity was flattened into prose on the way out, so the settings page could
only print the sentence (which also echoed the endpoint body unredacted), and
the refresh button invented its own literal error when a check came back empty.
Blast radius: /providers/check on failure now answers 200 with
`{ok:false, code, params?, httpStatus?, detail?}` and no `error`. `code` is the
dotted identity the add flow's refusals already use (provider.probe.*), plus a
new provider.probe.timeout split out of unreachable; `params` carries only the
numeric facts the sentence needs; `detail` is the endpoint's own text through
the same redaction the model check uses. Success responses are unchanged.
catalog.ProbeError gains Status and Body, read from the typed fetch error, and
the status/code table is shared by the refusal and the finding.
Neighbouring behaviours tested: refused, forbidden, payment, rate limit, missing
path, upstream error, no chat models, unreachable, timeout, success, key
redaction in the endpoint text, bounded one-line detail, missing grant,
unknown name, and the add flow's probe codes all translated.
Why this layer: the probe owns the class; serve only projects it; the page
maps the code to words.
The refine button's tooltip carried the Ctrl+Shift+E shortcut, but nothing
pointed the button at it: it had no aria-describedby and the tooltip had no
id, so a screen reader announced only "Optimize prompt" and the keyboard
route stayed undiscoverable. The attach button beside it already describes
itself this way.
The tooltip now takes an id from useId and the button names it in
aria-describedby, matching the attach button.
Refs #12156
A settled render_chart call (direct or through use_capability) is re-validated
on the page, lifted out of the folded execution group, and drawn as a bar,
line or pie chart from the validated spec by our own React SVG components in
a lazy chunk. A running, refused or invalid call keeps the ordinary tool card.
The same card is rebuilt on reload from the stored call. Each chart offers an
accessible data table, keyboard-reachable points and no entrance motion.
A settled chart call (direct or through use_capability) now draws its title,
marks and rows as a capped, width-fitted table in the terminal, with numbers
right-aligned and the fold key opening the rest, instead of a line count.
chartspec.FromCall and Spec.Table/Describe are the single projection; the TUI
history record and frame merge now keep the resolved tool name, without which
neither a reopened session nor a live call could be told from any other
use_capability call.
A successful save remounted the editor synchronously while the list reload
was still in flight, so the new editor was seeded from the stale entry and
showed the pre-save values. onEdited now returns the reload, the editor
waits for it before remounting, and save keeps the form busy until then.
A steer sent while a stop had paused the queue was refused with inbox.paused
but its queued copy stayed behind, so retries piled up and all ran on resume.
Hold the item as a follow-up under the admission lock and answer with the
queued receipt that already carries paused=true. The pause itself is kept,
and send-now from the queue panel still refuses. The composer reads the
receipt, from the steer path and from an idle skill send, and says the queue
is paused, once, in the notice strip.
Part of #12153
Stopping a turn while the plan approval card is open left the plan in
AwaitingApproval with no card to answer it. Later messages then ran under
the planning barrier and no new card could open. Treat a cancelled or timed
out approval as a revise so the workflow can propose again.
Part of #12153
The refine and attach tooltips are children of their buttons. They inherited
the button's white-space: nowrap, so the hint spilled past the 220px box in
both locales, and the disabled refine button dimmed itself to 35% opacity,
which multiplied down onto the tooltip that explains the control.
The shared .studio-control-tip now wraps inside its max-width, and a disabled
host dims its glyph instead of itself.
Part of #12148
Cause: the settings page awaited setRole and only then re-read /roles. While a
turn, a question or a background job is active the kernel has already written
the value and answers 409 runtime.saved_while_running; run() treated that as a
failure, the re-read never ran, and the controlled select snapped back to
"follow the main model" although config.toml held the saved model. The banner
also called it "操作未完成".
Fix: the role row always re-reads /roles (promise finally), and the page's one
failure banner now distinguishes saved-but-not-applied (SAVED_NOT_APPLIED, the
set the provider edit form already uses) from a real failure, with the same
title, warn colour and status role.
Blast radius: Settings.tsx only. Every action routed through run() now shows
"已保存，尚未生效" instead of "操作未完成" when the kernel answers one of the
SAVED_NOT_APPLIED codes; other failures render as before. No kernel change.
Neighbouring behaviours tested: role saved while idle, role saved while busy
(value and notice), real refusal (alert, no saved notice), other role rows and
a reopened page after a busy save, default-model row refused while busy (stays
on the kernel's default; the kernel refuses before writing, so it is not the
same defect). Go: POST /roles mid-turn returns 409 runtime.saved_while_running
and GET /roles returns the saved value.
Why this layer: the page owns the re-read and the banner; the code identity
already comes from the kernel and the shared SAVED_NOT_APPLIED list.
Part of #12144.
Cause
- `/quit` and `/exit` are in the slash catalogue but the TUI had no handler,
so since #11969 they were sent to the model as ordinary messages.
- Every Ctrl+C on a running turn re-sent cancel, so a turn that would not
stop could not be left, while the footer says "Ctrl+C exits".
- Tab in the completion menu moved the highlight down (and did nothing
visible with one item); the guides say Tab / Enter accept.
Blast radius
- internal/frontend/tui: send(), menuKey, the Ctrl+C branches of onKey and
viKey, one new Options field. internal/frontend/cli passes the quit names
from the slash registry. docs/GUIDE.md and GUIDE.zh-CN.md rows for Ctrl+C,
Ctrl+D and /quit. No controller, wire or config change. The bare words
exit, quit and :q stay ordinary prompts.
- Tab history: the cycling was part of the first completion commit
(5ff9d7a0f) with no stated reason and no test; nothing later touched it.
Neighbouring behaviours tested
- Idle double Ctrl+C, Ctrl+D (idle and while running), double Esc opens
rewind, a second Esc does not quit, Esc then Ctrl+C quits, one cancel call
per first Ctrl+C, vi mode Ctrl+C, a finished cancelled turn needs two
presses again, Up/Down still move the menu, Enter accepts, Tab accepts
the highlighted row, file-mention directory descends, Tab without a menu
still asks the kernel, /quit while a turn runs is not queued.
Why this layer
- Quitting is a frontend lifecycle action with no controller verb, so it
lives in the TUI local-command switch beside /mouse; the names come from
the registry through Options rather than a second alias list. Both
interrupt paths share one helper so Esc and Ctrl+C agree on what
"stopping" means. The unreachable shift+tab case in menuKey is removed.
When the kernel refuses browser.open, the workbench stored the reason in
`failed` but only drew it inside the file-tab branch. Opening the column
or pressing + with no page left the pane on its empty state, "choose a
file on the right, or open a browser", with no error.
Cause: `failed` was rendered only when the active surface is a file, while
the browser open paths (column opening with no agent tab, the + button)
also write it. The kernel also answered every failure with the generic
browser.open_failed, whose wording needs an {error} param it never sent,
so the typed engine_missing / engine_failed / profile_busy causes were
lost. Separately, `.workbench-error` used --danger, a token no stylesheet
defines, so such errors drew in the body colour.
Blast radius: WorkbenchPanel, one CSS declaration, serve's browser open
refusal and the kernel.ts / en_kernel.ts tables. The alert renders for any
non-file surface; file tabs keep their own copy and are unchanged.
`failed` is cleared when a tab is picked and when the active tab closes.
POST /browser/open now answers browser.engine_missing, browser.engine_failed
or browser.profile_busy for those causes (same 400, same params, plus the
detail in params.error) and browser.open_failed for the rest. The colour
fix also recolours the file-tab and explorer errors.
Neighbouring behaviours tested: the three codes and the generic one through
the real SsePort decoding path, on both open paths, asserting the exact
text and no leftover placeholder; the reason clearing on picking a file
and on a later successful open; a file read error staying single on its
tab and going away with it; serve end to end with real launch failures;
the refusal-code parity test; the whole WorkbenchPanel suite and tsc.
Why this layer: the pane owns the state and the draw; the kernel boundary
in serve is where the typed cause becomes a code, and internal/platform/
browser is not touched.
Fixes#12139
Add internal/base/scratch: a temp directory claimed by a long-lived
exclusive .owner.lock (internal/base/filelock). Remove releases the lock
first and retries sharing violations for up to 2 s. Sweep/SweepOnce remove
only non-symlink directories past a one-minute grace whose lock can be
taken, so a live owner is never touched. The jobs manager now creates its
reasonix-jobs-* fallback root through it, which also sweeps roots left by
earlier killed kernels.
Part of #12142
In 1.x, "/resume <n>" switches straight to the nth saved session. The Studio
TUI ignored the number and always opened the picker, the known gap left in
the CLI migration tracking issue.
An indexed /resume now resolves the session by its row in the same list the
picker numbers and resumes it, with the same refusals the picker gives: a
running turn, the session already open, or an index outside the list. A bare
/resume (or a non-numeric argument) still opens the picker.
The index lookup shared by /takeover <n> and /rename <n> <title> moves into
one helper so the three commands cannot drift apart.
Refs #11493
The feedback dialog is fixed to a 720px column and thread bubbles are capped
at 560px, so a long maintainer reply on a wide window reads as a narrow strip
that needs constant scrolling.
Add a "Fill the window" toggle next to Close. While pressed, the dialog takes
the whole veil and thread bubbles widen to 92% of the column; pressing it
again returns to the reading column. The toggle is a view-only action
(feedback.expand) and the state lives only for the open dialog.
Refs #11575
The reference-solution half covered only the e2e, train and verification-stress
corpora, so three committed corpora with the same tasks/ layout carried no
evidence that any attempt can score: a grader that can never pass is
indistinguishable from a task with no solution, and a corpus of those teaches
that nothing is ever accepted.
Their pristine seeds already fail, checked across all eleven tasks, so this adds
the reference solutions the guard needs: one ANSWER.md each for the three fleet
tasks, and the non-positive-skipping calc.py the eight project-check graders
assert. The e2e suite stays the single declared exception, named in the skip.
Refs #11852.
Prompt refine and commit-message drafting built their provider with effort forced to "disabled". Generic OpenAI-compatible endpoints refuse that at construction, so the side model was nil and the user saw "no usable model" while chat worked. sideProvider retries only on provider.ErrEffortRefused, with the lowest effort the entry's vocabulary lists (automatic depth when it lists none); other construction errors are not retried and are logged at warn.
When a summary fails under the pressure trigger the retry is held, and
while it is held and compaction is due nothing reached the user. The
agent now emits a coded notice, compact_held, once per turn and cause,
with the failure's code as detail. The desktop and the TUI word it from
the code in the UI language; a code a build cannot word keeps the
kernel's English.
A transient summary failure held the retry until the input grew by an
eighth of the trigger. Recorded near 105% of the trigger that release
point is the window itself, so the next request went out uncompacted.
The step is now capped at half of the room left below the hard ceiling,
so the hold always lifts before it.
A relay that refuses an oversized request with a 400 and no error code
was not recognised as an overflow, so recovery never ran. A 400 or 413
whose request our own estimate puts over the window now takes the same
fold-and-replay path, judged by status and size, never by wording.
A session row was selected only once POST /runtimes (or /resume), GET /runtimes
and GET /tree had all returned, so every switch to a session without a pane
waited out the kernel's pane assembly with nothing but a dimmed row to show for it.
The window now owns which session is being opened (useSessionSwitch). The row
reads selected and busy from that state as soon as it is clicked, the active
pane shows aria-busy, a ticket orders concurrent opens so the last click wins
and a refusal from a superseded open changes nothing, and a failed open drops
the selection back to the pane that is still in front while the error is shown.
Opening moves out of App into useOpenPane so every caller shares the one path.
The pane burst after an open now asks for /runtimes and /tree through one
single-flight read instead of one per caller.
/status called config.Load() on every poll, which re-parses every
installed plugin package under a process-wide lock. The controller now
carries the model entry boot resolved and /status reads effort, vision
and the model ref from it. The provider label still comes from the user
file because a rename rebuilds nothing.
Refs #12055
Discovery read every directory-layout skill's references/ and scripts/
to build a Body that a catalog never uses, so opening a pane cost
O(bytes in the skill library). Listing now keeps only SKILL.md; Read,
ReadSlash, Prepare and the capability digest complete the skill when it
is invoked or inspected.
Refs #12055
A summary request was cut off at a fixed 90 s of wall time, so a local model
that was still streaming steadily failed every /compress with summary_timeout.
Bound the silence between streamed chunks instead (6 minutes, one minute past
the providers' own 300 s stream idle watchdog), and add a 30 minute overall
ceiling with its own typed code summary_ceiling so a trickling model cannot
hold the turn. Both failures stay transient: a later attempt may retry once
the input has grown. The notices state which bound fired and its length.
GET /skills resolved each skill's activation through Controller.SkillEnabled,
and every call re-ran config.Load(), which walks all installed plugin
packages. Resolve activation once per request via Controller.SkillActivation.
Refs #12055, #12051
Go's default User-Agent is answered with 403 by edge bot protection on
datacenter networks. Catalog and manifest reads now use the Studio update
identity shared with artifact downloads, the relay websocket handshake uses
the account client's identity, and feedback, telemetry, crash and account
requests fall back to the shared client identity when none is set.
The pane polled /status every 250ms with no in-flight guard. When the kernel
answered in 13-16s the unanswered reads filled the browser's per-origin
connection pool and the approve, plan-decision and answer POSTs queued behind
them.
Idempotent reads now share one request per path with a single trailing
re-read, the pane schedules its next poll when the previous read settles, and
decision submits get a bounded wait that fails with a typed KernelBusyError.
The retry status line showed "connection broke before the response headers"
for every retried failure, including 502/429 statuses, did not say what the
attempt was waiting for or when it counts as unanswered, and its seconds
counter ran from the first notice of the whole stall instead of per attempt.
The retrying event now carries a typed failure class (connection_closed,
timeout, upstream_status with the HTTP status, stream_idle, upstream_error),
the backoff before the next attempt, and the header wait of the coming
attempt. The producers set them: SendWithRetry knows the class, the adapters
pass their idle_timeout_seconds window, and the stream replay path maps the
stream-interrupt reason. The window words each class and phase (backoff
countdown, then a wait that restarts per attempt) from those fields, never
from text. The CLI machine record carries the same fields.
Provider settings gain one advanced row for idle_timeout_seconds (1..32767,
empty is the default), validated in the form and by the kernel's provider
edit endpoint.
Fixes#12046
The host book has no per-machine state: every entry is dialed on demand and every entry is
listed in the rail. A machine that is temporarily off the network, or one kept for the
occasional job, cannot be marked as such - the only lever is removing it, which takes its
folders and its forward rules with it.
- RemoteHostEntry gains disabled, written as "disabled" rather than "enabled" so an entry
written before the field existed still means what it always did, plus IsEnabled(); the
renderer writes the key only when it is set.
- One switch per row decides whether the machine is dialed, and that is all it decides:
live panes keep running and the rail still lists the machine, because a machine being
off the network says nothing about what this window should show. Whether this window
lists it stays a separate preference, and it is an immediate row action next to the
row's other actions rather than a control inside the edit form, which saves on submit.
- remoteHostDialable is the single answer to whether a machine may be reached: POST
/remotes/open refuses a disabled one with remote.disabled before dialing, and remoteLink
(the tree and directory reads, and removing a far session) asks the same question, so a
disabled machine is not reached for at all. A book that cannot be read is a refusal
rather than a dial.
- GET /remotes reports disabled and the edit endpoint carries it.
Project every string and list of an install plan through one bounded
projection shared by all frontends: names, paths, commands and arguments
render hidden characters as visible escapes, prose has them removed, and
each field, list and the action count has a cap. A cut ends the field with
an ellipsis and sets previewTruncated on the action and the plan. The planId
still digests the complete unprojected plan, and the market ledger records
the unprojected paths. Installed-package detail views (inventory, record,
runtime, warnings) use the same limits from the shared text utility.
#11897 stopped a session's model switch from rewriting default_model, which left the
settings row reading status.modelRef - the model the session in front is running - so
switching models inside a conversation still moved that row while the file it claims to
describe did not change.
The row now shows the entry the model catalogue marks as default, falling back to the
session's model when the catalogue carries no default, and the column beside it follows
the same value. `main` stays the session's model for the vendor-protocol inference this
component does.
Tested: with a catalogue default the row shows it whatever the session runs; without one
it falls back to the session's model; the settings-catalogue, ModelUsage and i18n gates.
The interface scale was declared four ways: the config comment said
0.8..1.6, the save clamped to 0.7..1.8, the slider ran 0.8..2.5 and the
presets covered 0.9..1.3. Dragging the slider to 2.0 showed 2.0 while the
kernel stored 1.8.
config.ZoomMin/ZoomMax/ZoomStep are now the only declaration. The save
clamp, the rendered config comment and the range /appearance announces
(mirrored to the page by wire-parity) all read them, and the slider is
drawn from the announced range instead of carrying its own. A stored
value outside the range reads as the nearest bound but is not rewritten
unless the user changes the zoom.
Ctrl/Cmd + '=' / '+' / '-' / '0' (read by key or by key position, so
shifted and non-QWERTY layouts work) step the interface scale through the
named presets (range ends outside them) and reset to standard, saving
through the same appearance setting as the Appearance control. Chromium
page zoom is not used.
The range is 0.8..1.8: 1.8 was the largest value the old save could
store, so nobody loses headroom. The keys exist only inside the desktop
shell; a browser tab keeps its own page zoom.
Fixes#12032
Part of #11962
A stable studio-vX.Y.Z that is the highest stable studio tag is created with
--latest and without --prerelease. Candidates (-preview.N, -rc.N) keep
--prerelease --latest=false, and recovering an older stable tag does not take
latest back. The decision lives in scripts/studio-release-flags.sh and reads
the tag grammar, not wording.
The GitHub release no longer carries a root latest.json. The 1.x desktop
updaters fall back to releases/latest/download/latest.json, and builds up to
1.18 do not validate the manifest and share Studio's minisign key, so a
Studio manifest there would offer them a Studio package. The file stays in
the R2 mirror under <tag>/, which is the only place Studio reads it. The
release now has 21 assets.
Adds a flags-script test and a workflow contract test, and updates the
runbook, READMEs, MIGRATING and ROADMAP.
A deny or ask rule such as Bash(rm:*) compared only the words the model wrote,
so `env X=1 rm`, `sudo rm`, `command rm`, `nohup rm`, `nice rm`, `time rm` and
`/bin/rm` were judged by fallback mode instead of the rule.
shellparse.PeelWrappers reads one simple command from the shell AST and
follows a fixed set of wrappers (env, sudo, doas, command, builtin, exec,
nohup, time, nice) down to the program they run, naming a path-spelled program
by its base name. Deny and ask matching test every layer; allow and session
rules are unchanged. A wrapper whose target cannot be named statically needs a
human instead of falling through to the mode.
apply() wrote every pack token, and the ink variables even with no pack,
as inline custom properties on <html>, so any author stylesheet needed
!important to win. The values now live in one rule of a <style
id="pack-theme"> created after the built-in sheets, written through the
CSSOM so a rejected value is dropped. The contrast tiers carry the inks in
the stylesheet, with --ghost on --faint's value, and perf/tiers.mjs holds
each tier to the legibility users already had.
A Claude Code Mod declares "modules" in hooks/hooks.json instead of a
"hooks" map. The importer returned early on an empty hook map, so the
modules were dropped silently and a Mod with skills previewed as full.
The model list collapsed entries by endpoint + model, so two providers
sharing a base URL with different API keys listed each model once and
the second account's refs vanished from the picker.
TestInstancesDifferBetweenRuns instantiated every task twice and failed
whenever two independent draws collided. Three tasks answer with a bounded
number (i70 quota 1200..9800, i47 lease 15..300, i112 lease 20..420), so a
run fails with probability 1/8601 + 1/286 + 1/401, about 0.61%.
liveRand now delegates to liveRandFrom(io.Reader). The test asserts the
invariant at the seed: two live generators do not share a stream, the same
entropy yields identical instances and different entropy yields different
values.
The walk tolerated only not-exist, but a directory being removed on Windows
fails to open with a sharing violation or access denied. Those errors now
count as vanished only when a re-stat shows the path gone.
Cause: the per-model check classified any 4xx it could not name as
unknown/rejected and the row printed only "request refused, not confirmed".
The kernel already carried httpStatus; the settings UI dropped it, and the
endpoint's own error text was never sent, so a local Ollama 404 (base URL
without /v1, model id without its tag), a 400 for a model that cannot take
tools, and a missing key on a System One entry all read identically.
Blast radius: POST /providers/check/model gains an optional `detail` field:
the endpoint's error text as one printable line. It is built in this order:
control characters and ANSI sequences replaced, the probe key masked in raw,
URL-escaped and JSON-escaped form (skipped for keys under 8 characters),
secrets.RedactCredentials, then cut to 300 runes. Folding and the key
replacement see the whole text; only the pattern scrub is bounded (8 KiB cut
on a word boundary, after folding). The model row shows an HTTP badge and the text inline. Status and
reason are unchanged and never derived from the text.
Neighbouring behaviours tested: Ollama-shaped upstream (correct, base without
/v1, tagless id, System One kind); 400/404/405/422 classification; detail
equals the upstream text, masks raw/escaped/plain keys, Bearer, ?token=,
Cookie, neutralises newlines/ANSI/NUL, cuts after redaction, cuts at the scan cap with the key split at every offset, with and
without leading whitespace, bounds a 4 MB body, leaves short keys alone (each mutation-checked); add and edit flows
share one fact helper; full ui vitest suite and the serve package.
Why this layer: the serve handler owns the typed result and the row is the only
renderer of it; classification stays where the failure is first known.
POST /browser/open answers 400 when the navigation fails even though the
window's tab already exists, and the link handler read every refusal as
"nothing opened" and sent the same address to the machine's browser.
The refusal now names the tab it left behind and the handler only falls
back when there is none.
A pane adopted at launch is not a folder anyone picked, so nothing listed it
once a later launch opened another one. Its conversations stayed on disk with
no sidebar row to open them from. The folder is now listed behind the existing
ones when the pane's first turn mints its session path.
A queued line taken back could fail with a spurious error on a repeated click,
leave a pending transcript row when its receipt landed after the withdrawal,
and reappear in the panel when an older queue pull answered last.
A truncated, timed-out or failed summary request blocked automatic
compaction for the whole generation, so context kept growing past the
fold trigger with no further attempt. Block release now reads the typed
failure code: transient request failures are released after the input
grows by one retry step; every other class stays blocked.
The digest request inherited the session's reasoning effort, so reasoning
tokens spent the same 16384-token cap the digest had to fit in and a
finish_reason=length dropped the whole summary. It now carries
EffortOverride=low; adapters apply it only when the endpoint's own effort
vocabulary admits it, so endpoints without a low tier keep their configured
depth.
Refs #11970
/compact notices were raw English sentences built from err.Error(), and an
aborted card with no code read "compaction done / folded nothing". Add
compacted/compact_declined/compact_failed notice codes (detail = the
compaction code), give cancellation, unclassified failures and a busy
session their own codes, and word them in the desktop and TUI catalogues.
Refs #11970
The terminal UI asked the kernel to refuse any slash line the registry does
not answer, so a typed path or half-remembered command dead-ended with
"unknown command". 1.x sends it as an ordinary message and says so (#5756).
The TUI client no longer sets refuseUnknownSlash; the controller's existing
default path runs the line through the normal turn pipeline. That notice is
now localised (en/zh/zh-TW) and carries the unknown_command code.
A built-in named first is that built-in whatever follows it, as in 1.x: the
kernel matches /context, /new and /clear by first word, and the TUI does the
same for /help, /clear, /setup, /version, /mouse, /resume and its display
commands, so none of them is sent as prose when given arguments.
Refs #11493, #5756
With no API key the connection panel opened silently and a typed message went
to the provider, which answered 401 with a pointer to a shell command. The
panel now names the model that lacks a key, and /submit refuses a turn for it
with provider.key_missing before anything is sent. The TUI returns the line to
the composer and points at /setup.
The usage range test assigned the process-global time.Local while goroutines from earlier tests still read it, so the race detector failed the package. usageWindow now delegates to usageWindowIn, which takes the zone trailing windows count days in; the test passes a fixed zone and never touches the global.
A provider whose natural key slot is taken gets REASONIX_CONNECTION_<id>_KEY. Collect wrote that provider and its key, but restore validation denied every REASONIX_ name, so the backup sealed fine and then failed to open as malformed.
Settings and feedback are answered by a pane's kernel. With no pane open, feedback silently did nothing and the settings refusal never cleared. Opening either now starts a session in the current folder, a refusal is shown for both, and it is withdrawn once a folder or pane exists. The cloud-backup button now states what it is waiting for.
Preserve ACP frame-order assertions while skipping only MCP status updates,
and use t.Context in the serve regression test.
Report servers requiring launch approval as pending rather than failed in
controller, ACP and /mcp snapshots. Keep pending rows disabled without a retry
action while retaining their recorded transport and sanitized error.
Document connecting and awaiting-approval notifications, style the pending
indicator in both themes, and derive frontend state coverage from Go constants.
Add controller, endpoint, ACP and row-rendering regression coverage.
- Map the new `pending` state in the desktop panel with an i18n label. A
repository-declared server nobody has answered for is off, but not because
the user switched it off, and the row rendered the raw state string.
- Announce what actually happened. RecordFailure and
RecordLaunchApprovalRequired now share one store path and name their own
state, so a project MCP awaiting authorization stops reporting a failed
connection.
- Correct the MCPSurfaceReady contract: it fires on every change to what /mcp
answers, not once per server. Audited every consumer before changing it —
the ACP sink republishes a full snapshot, the serve broadcaster numbers the
frame and the SPA refetches /mcp, the headless stream passes it through — so
repeats are harmless and the documentation was the stale part.
- Keep the spawn claim beside the rest of the deferred-spawn bookkeeping
instead of a file of its own.
- Add the /mcp endpoint regression test the refactor was missing (live
transport and source, recorded status code, in-flight spawn, cached
description and tools), and a panel test tying every state label to the
English catalogue.
SetCredential and SetCredentialIfRevision built a KEY=value line and parsed it
with the dotenv reader, so ${...} was expanded and wrapping quotes stripped
before the key was ever written. The writer also left a value containing $
bare, which the reader then expanded. Store the assignment directly and pick
the first encoding the reader returns unchanged; a value no encoding can hold
(for example a trailing backslash next to a quote or $) is refused before
anything is written, so one entry can never make the whole file unreadable.
Adds GET/POST /display-currency over the controller's new
DisplayCurrencySettings/SaveDisplayCurrency (shared with /currency), a
Settings > Usage control, and makes the quote sink follow the
display_currency notice so rounds after a change are priced in it.
Picking a 1.x data folder only looked at <root>/sessions, so per-project
sessions, sessions-v4 and desktop-sessions-v5/by-id imported nothing and the
UI reported no old sessions.
render_chart validates a chartspec v1 spec with chartspec.Parse and answers
with a chart_id and the deterministic text summary, never the rows. It is a
registered built-in left out of the provider allow list, so it is found and
called through use_capability and the cache-stable prefix is unchanged. The
spec persists in the call arguments the session already stores (#11564).
cli-channels and cli-pointer both waited only for cli-tag, so the cask could be
pushed while the v<version> release holding its archives was still being
created, and stayed pointing at nothing if cli-pointer then failed. The cask
step also overwrote the tap whatever its version, so a workflow_dispatch of an
older tag moved Homebrew backwards; it now skips a tap that is ahead and fails
closed when the tap's version cannot be read.
Leaf package internal/contract/chartspec: spec types for bar/line/pie, a
strict Parse (size cap first, unknown fields refused, scalar cells only,
unsafe integers refused), typed refusals with dotted codes and caps, and a
deterministic Summary that quotes cell data. No consumer yet (#11564).
EditConn titled every outcome "not saved" at warn level, so a write that
landed but was not applied to the live session read as a failure, and a real
failure read as a soft notice. Split by outcome: error alert for a refused
save, status note for saved-not-applied, own title for a catalog refresh.
The archives that reasonix upgrade, the Homebrew cask and the GitHub release
install were linked with main.version only, so they reported no git_commit or
build_time_utc and the embedded product docs identified themselves as dev. The
npm packages of the same release were linked with all of them.
CostQuote carries a rate band for a rate confirmed as the vendor's own
schedule, aggregates fold it to peak, off-peak or mixed (none when any call
has no schedule), and the TUI prints it beside the cost as 1.x does.
A headless run in a folder with no trust decision refused every edit and
command and exited 0, so scripts that relied on 1.x defaults looked
successful. It now exits 4 (3 under --fail-on-unverified); the gate and
what it allows are unchanged. docs/MIGRATING.md gains the 1.x differences:
trust prompt, untrusted-folder runs, compact-ratio range, env vars.
Refs #11493
The panel dropped a list as soon as every item was completed, so the final
N/N never showed. It now stays through the ask panel and the idle prompt and
goes when the next turn begins; a rebuilt screen (resume, rewind) still does
not draw a list that finished before it was opened.
Adds a typed Commits port on the controller: ProposeCommit reads the staged
index through gitcmd, frozen as a tree object, and asks the session's model
for a conventional-commit message in a request of its own (prefix untouched,
billed as commit-message). CommitStaged writes the reviewed tree with
commit-tree and moves HEAD only if it has not moved, so an index that changes
after confirmation cannot enter the commit. Files named as secret holders or
whose added lines carry a private key are withheld from the model and need an
acknowledgement. Hooks and signing never run. Nothing pushes.
Refs #11518
- Add the pre-5.3 GLM catalogs to shippedPresets so installs holding the
old list are moved forward, plus upgrade tests (old shape upgrades, a
curated catalog stays untouched).
- Declare the per-model ladder once in provider.ZhipuEffortContract and
read it from the /effort menu, the normalizer and the request builder,
instead of repeating the model lists across three switches.
- Cite Zhipu's Core Parameters reference for the documented levels
(5.2's none/minimal/low/medium/xhigh folds, 5.3's forced thinking).
- Add a boot.Build effect test at the provider-request boundary covering
the stored enabled/disabled -> EffortDisplay -> request path.
- Surface the disabled->low landing for GLM-5.3 in the /effort menu and
document the 1M context window the added models reuse.
CodeQL's go/command-injection models regexp matches as barriers only when the
match is the branch condition dominating each use. screen() folded the match
into an || condition and the -C directory skipped it, so alert 1014 kept its
flows through screen() and Args.
- keep typed field details for days and explicit ranges
- reject from/to combined with days
- treat explicit dates as zone-less calendar days
- relocate and extend usage range tests
rebuildFailed collapsed the busy refusal into runtime.rebuild_failed with
the English sentence as detail, so the compaction, shell, sandbox,
permission, browser and MCP-load saves told a zh reader the write had
failed. It now answers runtime.saved_while_running, which the UI renders.
A bare https://host passed the add-provider probe because {base}/v1/models
answered, while chat posts to the saved base + /chat/completions and failed.
The probe now reports the base that was proven (typed address completed with
/v1 only when the listing answered there alone, openai-style kinds only), and
add, onboarding and edit-refresh adopt it into the address field before save.
F5 and Ctrl/Cmd+R reloaded the whole Studio UI from any focus, so a page
or browser-pane user pressing them lost the window state. The app window
now reloads only on Ctrl/Cmd+Shift+R (white-screen recovery, crash reload
unchanged); the browser pane reloads its own page on F5 and Ctrl/Cmd+R and
ignores the cache on the shifted chord.
POST /model persisted every switch as the user's default model, so changing
the model inside one conversation silently changed what every new session and
the next launch started on. The typed /model command, the terminal UI and ACP
never did this; only the HTTP picker path did.
The request now carries an explicit `default` flag. The Settings default-model
row and the first-run model choice send it; the composer picker does not, so
it changes only that session. A pane resolving through the hub's resolver
still never writes this machine's config.
One structural rule (secrets.RedactEndpoint / CredentialKey / RedactArgs)
replaces mcpsetup.RedactURL's query-only masking. Userinfo and fragments are
dropped or masked, query and fragment keys are decoded before classification,
unparseable endpoints render as <redacted>, and transport diagnostics keep
error identity while omitting peer-supplied prose.
Operational values (install, approval fingerprints, persistence) are untouched;
preview and plan responses add separate display fields that the desktop mirror
and confirmation UI render.
Rebased onto the launcher-aware NameFromArgv: names derive from the original
argv but never from a value the display projection masked, so build-flag values
such as `go run -exec` still resolve. Install classification reads the typed
HTTP status instead of sniffing the server name out of the message.
Fixes#11785
The primary button is focused from a passive effect that runs after the
element is committed, so asserting document.activeElement right after a
findBy* resolves races that effect. Use waitFor, as settings-layers does.
testenv.TempDir retries RemoveAll for two seconds when a file stays held.
On Windows the usage and history SQLite catalogs, which the shared helpers
left open, held a file under nearly every test's directory, so each of those
tests paid the full retry budget at teardown. Closing the catalogs from the
helper's own cleanup, which runs before the removal, makes it converge on
the first attempt.
A successful Windows bash probe is remembered per executable identity (path, size, mtime) for the life of the process; failures are never remembered. Every boot.Build used to spawn Git bash for it.
agent.Options.WorkspaceScanLimit (zero keeps the 50,000 production bound) is threaded from boot.Options. The incomplete-prose-scan effect test now sets it to 64 and writes 65 files instead of 50,001.
The deadline starts at launch, so with a real interpreter the test measured startup time. Under load Git bash needed more than 500 ms and the pre-deadline write never landed.
TestRememberPermissionRuleSerializesConcurrentWriters queued 32 writers on a
lock that waits 5s, so a slow Windows filesystem ran the last waiters out of
time (lock timeout at 5.08s). The cross-process variant had the same shape:
each helper waited a fixed 5s for the others to spawn. Eight writers still
interleave, and the readiness waits now read the shared hang budget.
The deadline is armed at launch, so a 500ms budget races interpreter startup:
on a loaded Windows runner PowerShell had not printed anything when it
fired, and the test reported the pre-deadline write as lost. PowerShell now
gets a 10s deadline named for what it has to outlast; POSIX shells keep 500ms.
Hang-catching waits in the control tests were hard-coded to 2, 3, 5 or 10
seconds. A whole turn through the real agent loop can exceed that on a
loaded Windows runner, so TestTurnOutcomeClassifiesPerseverationPause failed
with 'timed out waiting for TurnDone' on a run that passed on rerun. They now
read testenv.Budget(t), which is the single named bound for a wait that only
exists to report a hang.
cloudPresence kept a monotonic counter, so every reconnect of the same phone
took the next number. Allocate the lowest ordinal not held by a live
connection instead; it stays fixed for the connection's lifetime.
The wide-workspace effect test compared total goroutine growth against a
magic 64, which host background work (sqlite watchers in the history
catalog) pushed past under load. Count only goroutines inside the walkers
and bound them by pool size times concurrent walk kinds; assert none
survive the turn.
A fold that was refused or failed ended with an aborted card reading
"folded nothing" and a blocked receipt whose code was empty; the cause lived
only in a sentence. Checkpoint rejections now carry a code per rule, summary
failures are classified by sentinel (truncated, timeout, input too large,
context changed, provider failure), and the code rides both the receipt and
the aborted compaction frame. The Studio card shows a localized cause, and
keeps its before/after token pair on one line when the row wraps.
The Windows runner's temp dir is an 8.3 short name. The grant target for an
ambiguous path is resolved to the long form, so the expectation built from the
unresolved root never matched.
The regression held row and button references across the refetch that
PackageList.submit now triggers after a successful review submission, and
its deferred submitMarket fixtures never updated the inventory myMarket
returns. Query live rows after each reload and delegate to the mock's real
submitMarket so the reloaded inventory reflects the new status.
A tilde inside a Windows 8.3 short-name segment (RUNNER~1) of a drive-letter
path is an ordinary character; any other tilde can expand to a home
directory and stays refused as non-literal. Unescaped braces are refused
too, since a brace word expands to several targets.
Also give the PowerShell delete-analysis subprocess test a hang guard
instead of a 15 s latency assertion that fails on loaded runners.
The reference-solution guard fired only for the train corpus. For every other corpus a missing solution/ was a silent skip, so a grader that can never pass stayed indistinguishable from a task with no solution - the case the rule above it exists to prevent. #11380 supplied the ten vs-pos solutions; this extends the guard to the whole verification-stress corpus and supplies the ten tasks it was still skipping.
The e2e suite keeps the skip, and the skip message now says why: its handwritten tasks do not commit a reference solution.
Verified: go test ./cmd/e2ebench -run TestCorpusGradersPassTheReferenceSolution is green across e2e, verification-stress and train. Removing any single solution/ fails the guard at corpus_test.go:254 (checked by hand for vs-neu-cfgkey), so the added solutions are what the guard grades.
A relay that rejected the continuation field with a message the client did not recognise surfaced the HTTP 400 to the user, and the stored response id was kept so the next turn sent the same rejected request. Any HTTP 400 on a request that carried previous_response_id now retries once with the full history; a completed recovery stops sending the field for the client's lifetime, a failed one surfaces its own error naming the attempt.
Revision after the security review: ordinary commands with dynamic executable names, non-recursive deletes and non-delete wrappers launch again; each refusal carries its own cause and remedy; bounded compound sequences with literal targets inside the granted roots are allowed; default-parameter, pipeline, alias, .NET, find, xargs, cmd and static interpreter payload deletes are judged; symlink components are resolved before dot-dot cleaning; the session temp root is a granted scope; the PowerShell parser is reused across calls; bilingual copy for the new codes.
Doctor now validates declared agent roots fatally, which the warning test added by the root-diagnostics change did not expect, so the two merged changes failed together. The warnings are printed first and the root rejection keeps its exit status.
Approval alone did not stop a recursive delete: in unattended Auto and YOLO a PowerShell sequence whose first assignment failed non-terminally went on to delete the real home directory. A pre-launch gate now parses the command (bash AST, PowerShell parser run as data) and refuses recursive deletes whose target is the home directory, a drive or filesystem root, a workspace ancestor, an expansion, variable or glob, or outside the workspace and granted roots, with a typed shell.destructive_target code; parser failure is shell.analysis_unknown. The gate runs after approval and before every launch route.
A write to a markdown or text file in a workspace with no declared or captured checks still added a stale_verification debt and, in Balanced, a readiness floor demanding a verification command. When every proven mutation is a supporting-role path and nothing declares or captured a check, neither applies. Mixed, opaque, unproven or partially observed effects and declared checks keep their behaviour.
The sidebar row and the account panel disagreed when a token was present but the identity lookup failed: the row showed a signed-in state with a question-mark avatar while the panel drew the sign-in form and hid the error. Both now render the signed-in state, the identity error, a retry and confirmed sign-out.
The record panel was rendered after the whole active-turns list, so with a few
hundred rows it sat thousands of pixels below the click. It now renders inside
the clicked row's list item (or under the signal track for a signal), the same
item stays mounted, and clicking again closes it.
The session filter row gave four equal cells to labels with counts, so at the
old 176px floor the labels collided; the new-session and search rows wrapped
around their shortcut hints. Raise the column floor to 232px, let the filter
cells size to their labels, ellipsize labels, and drop counts and shortcut
hints below a container width.
The session and project menus wrote clientX/clientY into left/top, but the
interface scale is CSS zoom, so a fixed element's left/top is multiplied by the
zoom on its way in. Both now go through pinToViewport, which divides by the
zoom and clamps the measured box inside the viewport. The session menu's
max-height is divided by the zoom as well.
The official DeepSeek entry's model list migrates the retired
deepseek-v4-flash name to deepseek-flash, and qualified refs move with it,
but a bare default_model = "deepseek-v4-flash" (the shape a shared 1.x/Studio
config carries) was left behind. It then resolved to nothing, and because an
unnamed boot preserves an unknown default for its error, the Studio host's
only pane failed to build and the host exited 1 before its handshake.
- config: migrate a bare retired flash ref (default_model and every other
model-ref field) to the official entry's successor, unless another
provider still serves that name.
- boot: a new OpenOnFallbackModel option, set only by the Studio host and
hub panes, lets an unnamed build pass over an unknown non-plugin default
like a keyless one, open on the first configured chat model, and report
default_model_unavailable. The config file is not written. run, the TUI
and every other build keep the actionable unknown-model error and never
move to another provider on their own.
- serve: removing the provider in use moves the conversation to the startup
model when the saved default names nothing, instead of answering 500
after the removal; session titles follow the same startup model.
A background sub-agent keeps emitting reasonix.subagent.* progress after the
dispatching call has returned its job id. Folding those frames into the call's
card re-opened it, renamed it and its execution, and nothing closed it again.
Progress now lands in per-call phase state beside the cards; the Subagents
panel reads a child as running while its call is open or its last phase is not
terminal.
Take back cancelled the entry and dropped its text, so the label promised a
retrieval nothing performed. The body is now read before the entry is given
up and appended to the composer, under whatever is already typed. A line whose
body cannot be read stays queued rather than being lost.
The header chip counted client-side pending rows, which a delivery reported
before its HTTP receipt left pending forever, so it named a line the queue
panel (the only place to act on one) no longer showed. It now counts the
kernel's waiting guidance through the same predicate the panel filters by,
and a receipt that arrives after its delivery no longer re-opens the line.
The terminal UI opens a Configure connection panel when the session owes a
key, and /setup or /auth opens it at any time. The key is tested without
being stored, saved to the credential store only, and applied to the running
session when it belongs to the active connection; config.toml is untouched.
The Studio terminal UI answers the 1.x commands that read or steer the
kernel's own state, with 1.x argument shapes and output. Queue verbs and
steer go through the existing inbox routes, so the controller still owns
admission and ordering. /export is a controller capability: the file is
created exclusively and never through a symlink. /takeover resumes the
target without taking a session from another process.
foldTool looked for a call's parent only among top-level cards, so a
worker's own calls, whose parent is the per-worker task card nested in the
fleet group, fell through to the main list.
Pastes of five lines or a thousand characters become [Pasted text #N · M lines],
counted whatever row ending the terminal uses, and expand on send into the
labelled Begin/End block 1.x sends. Numbering continues past labels a resumed
session already holds, and panels receive pastes unfolded.
The footer's data row now leads with workspace@branch and the +added -removed ?untracked
counts from a new GET /workspace/git, read at start and after each turn. The compaction
meter label is back to the bare COMPACT word.
The one-time confirmation gate is gone: YOLO is always a stop on the cycle and a
single Ctrl+Y enters it, as in the 1.x terminal UI. --yolo on the command line
keeps its one-time notice, and the sandbox still applies in every mode.
On a fresh HTTP/2 connection, api.deepseek.com's edge acknowledges an eagerly
sent request body (about 8 KB or more) with WINDOW_UPDATE before its own
SETTINGS frame. RFC 9113 §3.4 requires SETTINGS first, so Go's client aborts
with "received WindowUpdateFrame before a SETTINGS frame" and the turn fails
with connection error PROTOCOL_ERROR. Studio sends a ~30 KB body on every
turn, so every request fails on affected networks.
Send Expect: 100-continue on chat requests to the official DeepSeek host.
The transport then holds the body until the server answers, by which point
its preface has arrived. HTTP/2 stays negotiated and request bytes are
unchanged; other providers are unaffected.
Moving the header setup into chat_headers.go keeps openai.go within its
file-size budget.
Refs #11679
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Next used the first still-open question, so after going back it skipped answered questions, and with two questions it vanished in favour of Confirm. Stepping is now positional; only the last question looks for an open one.
The workspace label in the sidebar footer, the composer's model name and the
workbench file name paired overflow clipping with line-height 1, so a UI font
whose ascent and descent exceed 1em lost the tops and bottoms of its glyphs.
They now use line-height 1.5; the rows centre their children, so nothing moves.
A style contract fails any new rule that clips a line tighter than 1.2.
The rail items re-declare the font shorthand, which resets
font-variant-numeric, so the digits were proportional and each item changed
width with its value. Restate tabular-nums in those rules and reserve the
width of a three-digit speed.
A submit that gets an answer without a recognised service code (an uncoded
5xx or an undecodable success body) is retried once with the existing
backoff, keeping the idempotency key. Coded refusals, refused redirects and
replies are never retried. The failure message now says the typed text is
kept, to try again shortly, and points at the GitHub alternative.
The source-name field intentionally hides validity errors while an IME candidate is still composing. The Add button did not share that state, so a completed form could submit a partial non-ASCII candidate before composition ended. Hold the button while composing and cover the name boundaries at the UI and serve boundary.
Probe host capabilities before opening a native picker. Headless and remote browser clients now enter a server-side path and use the existing /tree/workspaces API, while the desktop shell keeps its native dialog.
Archive a folded legacy recovery row together with its siblings, keep siblings open in a pane active, and restore sidecars if a multi-file update fails. Refs #11581.
The context-budget notice reported "room remaining" and, at the second
rung, told the model to land what it knows now. It never said that the
fold is automatic and the same task goes on, so a model near the trigger
read the figure as the end of its window, announced it would be
compacted, and handed the work back to the user (in-app feedback
FB-J2QR-HQ89).
Both rungs now state that the figure counts to the automatic compaction,
not to the end of the window, and carry one shared continuation clause:
keep working, do not stop, hand the work back or suggest a new session.
The notice still rides the turn tail and the typed ContextBudgetFigures
event is unchanged.
Two fixtures sat within a few tokens of the notice size and are widened.
Every runtime build ran ReconcileCleanupPending inline: a handful of stat and
read calls per transcript in the session directory. On a Windows box with 8000
sessions the sweep took 11 s of an 11.8 s build and 29 s at 20000 sessions
(about 1.4 ms per session, warm), and it sits before the host handshake, so a
cold start with a large history waits on it. The shell gave up at 20 s.
The Studio host, hub.Open, workspace switch and rebuild paths now hand boot a
reconciler that runs the sweep in a goroutine, once per directory at a time.
8000 sessions: build 11.8 s -> 1.0 s.
Refs #11434
Skipping an ask ends the turn through the same cancel as the stop button, so
guidance the user had queued into it was held uncertain and the queue paused.
The cancel now carries a typed cause: only the user declining a question
returns the undelivered item to the queue as an ordinary follow-up and lets
dispatch continue. Stop, errors and crashes keep the pause, and a stop that
lands after a skip still wins. The requeue only moves an accepted steer, so a
repeat delivers nothing twice.
Refs #11524
The add-source form only learned that a name such as a Chinese one is not
allowed after the kernel refused the save, and then showed the refusal under
the title "cannot connect". The name rule now shows inline as the name is
typed (with the allowed characters stated), the add button waits for a usable
name, coded name refusals from the kernel land on the field, and any other
failed save is titled as a save failure. The identifier rule is unchanged.
Reported through in-app feedback FB-WJ8B-C1XH.
Adds a user-level [auto_archive] setting (enabled = false, days = 30 by
default; a project reasonix.toml cannot set it). While on, the controller's
ArchiveInactiveSessions archives conversations idle past the threshold from
the same background pass that sweeps recovery branches, so it stays off the
boot path. Archiving only flips catalog metadata and is reversible.
Never archived: pinned, leased by a live run (covers a run or a pending
answer/approval), carrying an unfinished turn, the controller's own
session, subagent traces and superseded versions. Restoring a conversation
counts as activity so it is not re-archived straight away.
The kernel is the authority for pins (BranchMeta.Pinned, tree rows carry
`pinned`). The window sends the union of its own pins and the kernel's
after the tree loads and retries on failure; the sweep, including the one
after enabling, refuses to run until a window has synced once.
Archived conversations stay findable: search returns them with an
"archived" marker, and rows archived by the sweep carry a mark so the
window says once how many were archived and that they can be restored.
Settings > Session gets GET/POST /auto-archive with persistence to the
user file and zh/en strings.
Refs #11516
An optional user-level [serve] share_port makes the direct link listen on a
chosen port instead of a random one. Unset keeps today's behaviour. The range is
1024-65535, a project reasonix.toml cannot set it, and the bind range is
unchanged. A taken port is a typed refusal (share.port_in_use). The settings
page saves it to the user config; changing it moves an open door and
phones re-pair.
Refs #11540
The onboarding stage only had a min-height, so it grew with its content and
never overflowed itself while the body clipped it. Once the connection result
was showing, the Start button sat below the window with nothing to scroll.
The stage is now the window's height and scrolls; the shell no longer shrinks
or acts as a scroll container. Adds a browser guard (perf/onboarding.mjs).
A model told a user that Reasonix has no hooks. The guide's hooks section only
covered SessionStart, and the bundled reasonix-guide skill had no settings
schema, exit-code convention or worked example.
A ```diff fence rendered through the colourised built-in path on every
frame, and a colourised row costs ~1.4 ms to lay out (chroma via
regexp2) against ~1.3 µs for a plain one. While a fence streams every
delta re-renders the whole growing body, so a multifile diff pegged a
core and the display crawled. Changes:
- Each "@@ " hunk header settles the rows before it, so render that
prefix and leave only the in-progress hunk on the plain rail: the diff
colours in as it streams instead of only at the closing fence. A
closed fence settles the whole body. The formatter runs on each
settled prefix too, so its output appears hunk by hunk.
- Keep the already-formatted rows on screen while a settled body's run
is pending: growing the body by one hunk changed its key and redrew
the whole section on the plain rail until the new run landed, so the
delta output flashed back to plain at every hunk. The renderer now
draws the largest settled prefix whose result has landed and leaves
only the newly settled remainder (and the in-progress hunk) on the
rail.
- Memoize the colourised rows per hunk (hunk, path, width, rail, theme).
A completed hunk never changes, so it is highlighted once however many
frames redraw it; caching the whole body instead would re-highlight
every earlier hunk at each "@@ " and is quadratic in the hunk count
(measured: 60 hunks, 79 KB took 40 s with a body memo, 2 s per-hunk).
The theme is in the key so a runtime theme switch cannot serve the old
palette's colours.
- A fence still with fewer than two hunks keeps the plain rail, and a
closed fence whose configured [cli].diff_formatter result is still
pending also draws the rail, since that output supersedes the built-in
rows anyway; a landed-but-failed run falls back to the built-in.
- Hold the transcript viewport still when the content shrinks: a settled
diff collapsing its raw preamble to one formatted line frees rows at
the bottom, and following them back up made the screen jump. The
position now holds, the freed rows stay blank, and later rows fill
them, so scrolling resumes only when the output actually grows past
what is shown. The hold has only a lower bound, so a shrink larger
than the viewport would hold the position past the content and blank
the whole transcript — a collapsing multifile fence, or /clear
restarting over a shorter one — so the view re-anchors to its tail
instead. A manual scroll or resize releases the hold.
- Coalesce the stream frames the TUI hands the model: waitUpdate drains
everything already queued into one updateMsg, so a burst of deltas
costs one render instead of one per delta. Order is kept and a Gap
still triggers a history reload.
- Fix the file-header misdetection: a removed SQL "-- x" line renders
"--- x", which trimDiffPreamble/diffSectionStart mistook for a
"--- "/"+++ " file header, dropping the "@@" hunk header (line numbers
reset to 0) or the rows themselves when a removed "--- x" sat before
an added "+++ y". A pair is a file header only when a "@@ " hunk
follows it, and countDiff no longer skips "--- "/"+++ " rows.
- Keep a streaming multi-file diff's next file header on the plain rail:
splitAtLastHunk settled everything before the last "@@ " header, which
included the next file's "--- "/"+++ " pair (and its git preamble) once
that file's first hunk header had arrived but no hunk of it had
completed. isFileHeaderPair then saw no "@@ " after the pair, so the
settled renderer read it as a removed/added row of the previous file's
last hunk and counted it in that file's stat. A trailing dangling file
header now moves from the settled part into the tail until its "@@ "
arrives.
Tests: the multi-file cases TestSplitAtLastHunk was missing and an
end-to-end streamed test; a theme-switch test pinning the per-hunk memo's
theme key; and a benchmark replaying a 4-file x 15-hunk stream, redrawn
per streamed line.
fix(cli): highlight a streaming diff fence's hunk once
splitDiffSections appends a newline to every section, so splitHunks
returns a section's last hunk with a trailing newline and every earlier
hunk without. hunkRowsCached keyed the memo on the raw hunk, so one hunk
carried two keys over a stream: it was highlighted once while it was its
section's last hunk and again once a following hunk arrived. Every hunk
but a file's last was therefore highlighted twice, at ~1.4 ms a row.
Trim the trailing newline before keying. diffBody already trims, so the
rendered rows are unchanged; the streaming benchmark over a 4-file,
30-hunk, 1816-line diff drops from 7.8 s to 5.6 s.
TestStreamingDiffHunkMemoisedOnce pins one memo entry per hunk: it sees
three entries for two hunks before this change, two after.
Fold the writer-tool preview before highlighting it too. diffBody laid
out every row and then dropped all but maxLines, so a write_file card's
~24-row preview colourised the whole diff: ~772 ms per repaint for a
199-line file, paid on every message while the call runs. Counting the
fold from the source and laying out only the kept rows prices a
2000-line diff like a 24-line one.
BenchmarkDiffBlockPreview prices both shapes through a diffPreviewFold
seam, which TestDiffBodyFoldSeamSameRows proves draws the same rows: a
2000-line preview is 6.3 s and 86 MB with the whole body laid out, 71 ms
and 1.0 MB folded, and flat from 24 lines up.
docs(cli): correct diff-fence comment coherence
Four comments described behaviour the code does not have:
- diffRowCount's doc counted one row per non-blank line, but diffBody draws no
row for a "@@ " header, so it over-counted by the hunk count.
- renderCodeRail's doc claimed the in-progress hunk and the pending-formatter
placeholder, both drawn by renderRailText, not here.
- splitDiffSections' doc named only the fence path, omitting DiffText, and
described the plain-rail fallback where DiffText drops a headerless section.
- the file header claimed md.go still sits at the 800-line ceiling.
Reword each to the constraint that holds now.
A held queue item showed the kernel's English diagnostic (for example
"steer accepted but unapplied before turn exit") in a Chinese UI. Items now
also carry a stable BlockCode and the desktop words it in the UI language;
unknown codes fall back to the diagnostic.
A real failure that raced the stop button was reported as a cancellation and
hidden. TurnDone.Cancelled now requires Err to be nil or context.Canceled. The
desktop also reads "Cancelled" instead of "Interrupted" after the user's own
cancel.
Refs #11524
Settings > Storage moved sessions and projects but left other top-level state
behind while the code reads only the new root: the remembered project list
(serve-workspaces.json), user-level REASONIX/AGENTS/CLAUDE instructions,
memory/global, schedules, trusted seals, machine-id.key and the MCP migration
marker. All are now owned entries, so a move carries them.
Installs already moved get them back at boot: each entry is merged once without
overwriting and the old copy is renamed *.adopted. The project list is merged
under its lock and written atomically by a new leaf package shared with serve,
and file copies publish whole files. A guard test scans the tree for paths
joined under the state root so a future entry cannot be forgotten.
Both host notices reached every frontend as kernel English, so a Chinese UI
showed an English card. The context-budget notice now carries a code and a JSON
figures payload, the unapplied-guidance notice carries the user's text as
Detail, and the desktop card and the TUI word them from those in the UI
language. Unknown codes and undecodable payloads still fall back to the
kernel's text.
Refs #11524
Skipping an ask (or stopping a turn) ends it with the kernel's Cancelled flag
set and context.Canceled as Err. The transcript appended Err as an error card
regardless of the flag, so the user's own choice read as "context canceled".
The flag is the identity; Err is not drawn when it is set. An effect test pins
the kernel side through boot.Build.
Refs #11524
resolveWorkspace only knew a leading slash as rooted, so a drive-letter path such as C:\Users\u\x (what the remote tree reports as a project root) was appended to home and the serve never started. Rootedness is now asked of the target machine.
Headless `run` must leave stderr empty unless --debug is passed, so the
boot record of the resolved shell interpreter cannot be an Info log. It is a
diagnostic; the always-on counterpart is the Warn shellrun emits when a
launch actually fails.
Trim repolint comment blocks and move launch-failure logging out of RunForeground so the function stays within the 120-line limit. No behavior change.
Signed-off-by: Harbor404 <[email protected]>
A pinned [tools.shell] path was honoured on existence alone, so a pwsh.exe
that stats fine but will not start was held for the whole session while every
command failed at launch — the failure #10633 reported, still reachable
through the pinned-path branch after #10617 reworked auto-detection.
- Prove a pinned PowerShell the way the bash arm already proves its own, and
fall back to detection with a warning when it will not start.
- VerifyShell proves a pinned PowerShell too, so the settings screen refuses
one where it was typed rather than on every later command.
- powerShellLaunches answers false for a path that is not there: a missing
executable is not launchable, so callers that gate on it alone now agree.
- resolveShell takes the launch probe as an argument, which puts the decision
table ("pwsh will not start, 5.1 will") under test on any host.
- Detection that finds nothing says so instead of silently returning bare bash.
- Boot records the interpreter the session resolved to, and a launch-phase
spawn failure records the shell, its path, the caller and the error. The
desktop relays the sidecar's stderr into desktop-shell/logs/, which the
report found empty.
Windows end-to-end is not verified locally; the fallback is covered by the
injected probe on Linux and macOS.
Refresh the feedback unread count on window focus (at most every 30 s) and
every 10 minutes while a report is unsettled, with back-off to 1 h and no
error UI. A person with no open report is never polled. A hidden window keeps
polling only for someone who enabled the new feedback-reply notification.
Add a fourth notification kind, feedback_reply (default on under the master
switch), raised through POST /notifications/feedback-reply when the unread
count rises while the window is unfocused. The text never carries the reply.
The hand-written Edit menu dropped Paste and Match Style, Delete,
Substitutions and Speech; restore them with the same roles (structure checked
against the role-based menu in a real Electron). The menu factory is now
Electron-free and unit-tested, the language follows the first preferred system
language like the page, and the redundant second menu install is gone.
Refs #11524
The project menu offered only sort and remove-from-list. Add a reveal item that
asks the hub where a listed project lives (GET /host/workspaces/locate, window
only, refuses any folder the window does not list) and has the shell select it,
so it needs no open pane. Remote hosts draw in their own rail and never reach
this menu; a project whose folder is gone shows the item disabled with the
reason.
Refs #11524
Role-only menu items take their text from Electron's built-in English
catalogue even when the app locale is zh-CN (verified on a real Electron:
getLocale() == zh-CN, labels Undo/Cut/Copy/Paste/Select All). Keep the roles
for the clipboard work and give each item an explicit zh/en label chosen by the
page's saved language (rx-lang), falling back to the machine locale exactly as
the page does. The macOS Edit menu is rebuilt when the language preference
changes.
Refs #11524
Registers tool.show-all, agent.open, agent.close and agent.focus-trap, moves
focus into the overlay and back to the deck chip, caps the overlay's rendered
output at 200000 characters with a note, keeps the sub-agent count stable, wraps
nested output, and drops a stray .agents rule that changed the sidebar spacing.
Headless `run` must leave stderr empty unless --debug is passed, so the
boot record of the resolved shell interpreter cannot be an Info log. It is a
diagnostic; the always-on counterpart is the Warn shellrun emits when a
launch actually fails.
The sub-agent list stays after delegates finish and each row opens that
sub-agent's whole transcript. A nested call whose output exceeds the 400
character preview now says so with a show-all control instead of cutting.
Trim repolint comment blocks and move launch-failure logging out of RunForeground so the function stays within the 120-line limit. No behavior change.
Signed-off-by: Harbor404 <[email protected]>
Project configuration could replace a user server entry and drop disabled_tools, while a stale call to a filtered tool was reported as unknown. Merge user and project disables, reject empty names, and carry a typed refusal through boot and hot-added MCP registration.
Move an existing contextual-tool helper unchanged so execute_one.go does not grow past the file-size ratchet.
A pinned [tools.shell] path was honoured on existence alone, so a pwsh.exe
that stats fine but will not start was held for the whole session while every
command failed at launch — the failure #10633 reported, still reachable
through the pinned-path branch after #10617 reworked auto-detection.
- Prove a pinned PowerShell the way the bash arm already proves its own, and
fall back to detection with a warning when it will not start.
- VerifyShell proves a pinned PowerShell too, so the settings screen refuses
one where it was typed rather than on every later command.
- powerShellLaunches answers false for a path that is not there: a missing
executable is not launchable, so callers that gate on it alone now agree.
- resolveShell takes the launch probe as an argument, which puts the decision
table ("pwsh will not start, 5.1 will") under test on any host.
- Detection that finds nothing says so instead of silently returning bare bash.
- Boot records the interpreter the session resolved to, and a launch-phase
spawn failure records the shell, its path, the caller and the error. The
desktop relays the sidecar's stderr into desktop-shell/logs/, which the
report found empty.
Windows end-to-end is not verified locally; the fallback is covered by the
injected probe on Linux and macOS.
A run in a folder with no trust decision refused every edit and command
under the asking default and exited 0 with an empty stderr (plain text
run) and only the generic 'no interactive approver' reason for the model.
The refusal now carries permission.untrusted_folder. The model reads the
cause and the remedy, the text run prints refused tools, code and remedy on
stderr, the structured result's denial carries a remedy, and
--fail-on-unverified exits 3 for it. Exit 0 and the trust gate itself are
unchanged.
Refs #11493
The shell gave the kernel 20 s to handshake and reported every failure as a
sentence. The wait is now 60 s (REASONIX_STUDIO_HANDSHAKE_TIMEOUT_MS), a small
starting window appears after 3 s, an exit or spawn failure within 10 s is
retried once, and launch failures carry a typed code (timeout, exited, spawn,
overflow) with a matching explanation in the error box.
Refs #11434
Answering the last question of a multi-question ask with nothing left
unanswered commits the batch immediately, skipping the extra confirmation
Enter; an incomplete batch jumps to the first unanswered question instead. The
behaviour is gated on a new top-level `auto_submit` option (default false):
with it on the ask drops its Submit tab, and without it the Submit tab is shown
and still needs its explicit Enter.
`auto_submit` is a user preference: LoadForRoot holds it across the project
merge so a cloned repository cannot turn it on, and neither the annotated
project render nor the project delta carries it. Answering moves on to the next
unanswered question, so answering one the panel was thrown back to skips ahead
to the next gap rather than stopping on the already-answered question that
follows it; only the last question, with nothing left unanswered, commits. With
auto-submit on there is no Submit tab to page onto, so right stops at the last
question rather than landing on a hidden tab that commits.
- config: new top-level auto_submit, user/global only; rendered in the user's annotated TOML and omitted from a project render and the project delta.
- tui: Options.AutoSubmit; nextQuestion advances to the next unanswered question and commits on the last; right stops at the last question instead of a hidden Submit tab.
- program_test.go: the commit, next-unanswered-after-skip-back, skipped-question, hidden-tab, no-option and arrow-stops-at-last cases.
- docs: auto_submit in CONFIG_PATHS.md, reasonix.example.toml, and the Guide shortcut tables (en and zh-CN).
Move the providerSourceScope type out of config.go and the [ui] table render
out of render.go into their own files, so both sit under their recorded
repolint budgets with room for the options that follow.
My feedback shows the thread under each report as plain-text bubbles, a
needs-your-input chip with an inline reply box (byte counter, errors per code),
status chips for answered and closed, and a New mark that stays for the visit
in which the replies were read. The sidebar entry and the command palette item
carry an unread count that is read when the app opens and when the page opens,
never on a timer. The form copy says a report becomes public only if it is
filed as an issue, and keeps the warning about secrets and screenshots.
Refs #11424
The kernel reads the new statuses (needs_info, answered, closed), each report's
thread and needsInput, keeps the last-seen reply id per report locally and
exposes unread/hasNew for an entry badge. A reply goes through the controller
port, serve (POST /feedback/{receipt}/reply and /seen) and the TUI
(/feedback show, /feedback reply with --yes as the first token). The new
worker codes become sentinels: feedback.reply_limit, feedback.not_replyable
and feedback.challenge_required. A reply is never retried. Thread text is
cleaned of escape sequences before any frontend sees it. Submit copy no longer
says every report becomes public.
Refs #11424
The worker now holds new submissions by default. The converter still
creates issues only for received items, links only images the worker marks
released, and posts one idempotent bot comment per reporter reply on items
that have an issue, then acks it. Contact is never read by the converter,
which a test pins.
Refs #11424
A browser cannot read the status of a refused WebSocket upgrade, so a
rate limit, a refused origin, an expired ticket, a network loss and an
unreachable gateway all read as "relay temporarily unavailable". When the
socket errors before opening, one probe round (offline check, the gateway
health route, and an attachment preflight that only an allowed origin passes)
now classifies the failure by typed code; a 429 from the account service
carries its Retry-After. The notice names the cause, offers reconnect
(counting down a rate limit), and a reconnect that runs out of retries ends
with the last attempt's cause, or idle when the relay had reclaimed the
lease. The generic sentence remains only when the probe itself fails.
Adds internal/platform/feedback: env collection, local redaction (the
crashreport implementation, now exported), image re-encode with a metadata
self-check, idempotent submit with retry, persisted install identity and
receipts, and GET /v1/feedback/mine with typed sentinel errors.
The controller exposes it as the Feedback port (SessionAPI), serve answers
GET /feedback/env, GET /feedback/mine, POST /feedback and POST /feedback/name
with dotted refusal codes, and /feedback [list|name|<category> --yes <text>]
is a controller slash command so the TUI inherits it. Nothing reaches a model
request or the cache-stable prefix.
Refs #11424
Adds a scheduled converter (scripts/feedback-sync.mjs) that turns pending
Studio feedback into issues, idempotently via a hidden marker, and reports
in_progress/fixed/wontfix/duplicate back to the worker. A fix recorded as
"next" is upgraded to its studio-v tag once the tag exists.
exec.LookPath returns only the first bash.exe on PATH, which is normally the
WSL launcher in System32, and the fallback list covered only the default
Program Files roots. A Git for Windows install elsewhere (with just its cmd
directory on PATH) was therefore never detected, so Studio offered PowerShell
only. Candidates now also come from every PATH directory and from the Git
tree the git.exe on PATH belongs to, and the WSL launcher is rejected in the
candidate loops too.
Fixes#11421
Screened 87 candidates, staged: one pass over the pool, then two more only for
the candidates that were clean on both boundary arms. The six that held on both
arms in all three passes stay, the rest are dropped, so the efficiency substrate
only contains tasks that qualify:
quarter i62-epochlatch, i37-dispatchlag
half i106-budgetseal, i70-ingestfloor
default i47-tenantlease, i51-routedelay
All six are SnippetStop, SnippetThenRead or DefensiveRead in each of their
eighteen judged cells - nothing post-sufficient.
What the screening found: the answer kind decides whether a candidate holds.
Half-tier candidates whose answer was a bare number never survived (0 of 8);
the same tier with a codename answer survived 11 of 20. The dirty class in both
cases was PostSufficientRetrieval - the model re-checking a value it already had
- so the shape that needs no checking is the one that holds.
Also: i07 and i10 were re-measured after the review asked for it (their old
numbers came from the grep shape they no longer use), NeverSufficient is
reported separately rather than counted as clean, and the i11/i12 rows are gone
because those tasks are not in the corpus.
The host took its working directory as the first pane's project. A Start menu
launch inherits the install directory and a Finder launch the filesystem root,
so the app's folder was listed in the sidebar again after every removal. The
directory now counts only when it is outside the app's own directory and not a
filesystem root; otherwise the first remembered folder that still exists is
opened, then the home directory.
Fixes#11418
The provider form only edited one window and one output cap for every model
on a connection. model_overrides already carried both per model and the
kernel resolves them for requests and compaction, so the edit endpoint and
the form now read and write them.
A model with no entry inherits the connection's value, shown as a grey
placeholder; a blank field is never sent as an override, and clearing one
removes the entry.
A cli-pointer job publishes the Studio release's CLI archives as a GitHub
release vX.Y.Z (the only shape the 1.x updater accepts), records it
immutably in R2, then advances cli/{stable,preview}/latest.json. It reuses
the 1.x manifest validators and pointer ordering, only moves forward, and
is gated by STUDIO_PUBLISHES_CLI.
cli-gate requires CLI_PUBLISH_FROZEN so two lines never write the CLI
channels. cli-tag creates the vX.Y.Z tag with the release tag identity in
one inline step of a job that runs no repository code. An existing release
is trusted on its own checksums, never compared with a rebuild.
- The host creates the schedule store directory (0700) before a command's
sandbox is built, so the forbid-read mask exists for it on Linux and on macOS
paths that go through a symlink.
- Document in schedule/doc.go that the confirmed digest defends only against
sandboxed writers (Windows, sandbox off, yolo and MCP/plugin processes can
still write the store).
- Only the model's own usage report for an executor request counts toward
metering; estimates and side calls no longer let unmetered requests slip past.
- The job handle is released on the ready-refused, ready-timeout and cancelled
paths.
- A stored session path must be a .jsonl file in a sessions directory under the
state root.
- Tests: the report's untrusted mark, the stderr sanitising, the supervisor
watch that cancels and forces the process out, and bubblewrap tests for the
forbidden-directory mask. The child's own state check is dropped; Start
refuses a run that is not running.
Refs #11406
- A run whose provider reports no usage is stopped after two unreported
requests and charged the whole per-run cap instead of 0.
- The start of a run is one manifest transition guarded by a token that only
travels over the supervisor's pipe (its hash is what the manifest holds); the
child announces it holds the run's lease before it is released, and the
supervisor holds a supervision lock from before the spawn until the run is
settled, so the reaper never sees a gap.
- The read-only posture runs no config migration, so it rewrites no config file.
- Seatbelt denies writes as well as reads under forbid-read directories, as the
bubblewrap mask already did, so a blind overwrite of the schedule store fails.
- Windows refuses to run a child that a Job Object cannot hold; the job handle is
released; a child whose supervisor left is forced out after a grace period.
- The stored report is marked untrusted, the session path must lie under the
state root, the stderr tail is sanitized, the supervisor's backstops use the
policy minimum, the child reads the same user config file the loader does.
- Tests for the mutants the review found surviving; supervised runs are bounded
by a context so a missing kill fails instead of hanging.
Refs #11406
Adds the hidden `reasonix schedule exec <trigger-id>` child and a supervisor
that starts it, counts the tokens it reports and ends it, process tree and
all, when the wall clock or 1.25x the per-run token cap passes. A run that is
killed, crashes or loses its usage stream is charged the whole per-run cap.
The child holds an operating-system lease for the run, waits for the
supervisor's release line (so a run the store already settled spends nothing;
MarkRunning returning ErrRunSettled kills the child), reads its policy from the
user's configuration only, builds the read-only posture, and keeps reading its
stdin: the end of that stream is the supervisor going away, and the child then
stops and settles itself. A run that parks one kind of request too often
(max_repeat_parks, default 3, user level) is ended.
The final report and parked records are sanitized and bounded by the store on
the way in (PutResult). The schedule store directory joins the protected read
set of every run and the file tools' write guard; Grant.Tools narrows the read
tools of the posture.
Refs #11406
realPath no longer cleans a POSIX path lexically before resolving it, so a
'..' after a symlink is checked against the file the system opens. On Windows
it asks for the final path through a handle, which follows junctions and
settles short names, case and the extended and UNC spellings. read_file
re-checks the scope against the open file. LockSurface freezes SetTools and
SetExtensions once assembly has installed them; Sanitize now removes
zero-width, tag and line-separator characters and runs before a record reaches
any store; a workspace root is refused when the home directory is unknown.
Refs #11406
Read tools take readRoots, checked after symlinks resolve, with the typed
refusal workspace.read_outside_scope; walkers, glob matches and ripgrep
(--no-follow) honour it. The posture forces both secret-protection switches
on and refuses a workspace containing the home directory or filesystem root.
No hook loads under the posture (an empty non-nil scope list). Rebuilding an
observed controller without the posture is refused (ErrObserveRebuild) and
the executor's gate and asker are locked. Approval records keep the host's
text in Summary and the model's arguments in Detail, marked untrusted and
stripped of control and bidi characters. A run may park at most 32 records,
then it is ended (ErrParkLimit). A tool outside the ceiling is refused with
posture.tool_not_allowed instead of unknown tool.
Refs #11406
Adds a host-only Options.Observe that builds a run which can only read
files. The tool ceiling is decided by what each tool declares about its
reach (tool.Reach), enforced by a registry that refuses every other Add.
Approvals and ask questions are parked as Pending records and never
answered; the gate and asker cannot be replaced through the controller.
The workspace's reasonix.toml, .mcp.json, project hooks and folder grants
are not read, no MCP server or extension sidecar starts, and the run
context rides the turn tail as <scheduled-run>. The posture records
remote_content=false and whether an OS sandbox backs the tool filter.
Refs #11406
Slice 1 of the scheduled-runs P0 (#11406): a machine-wide manifest with atomic
writes and a schema version, a Policy that only the user layer can loosen,
pessimistic budget accounting, and claim-before-start with deterministic
trigger ids. No process is started and nothing is wired to a frontend.
The review settled the bar: a task qualifies only if it is clean on both of its
boundary arms, and none of the first six candidates is. This adds five more,
authored to the same shape but with a single value each and a call whose
arguments name the topic (./scripts/x {{cue}}), which is the shape the review
pointed at for i07 and i10.
- i15-ingest-lag and i16-commit-fence (quarter), i17-route-budget and
i18-salvage-epoch (half), i19-tenant-quota (default).
- All five preflight clean at their own tier's generation.
- The corpus stays unqualified until six candidates hold on both arms across a
three-pass batch; the ones that do not are dropped rather than promoted.
The title button in a memory row was flex:none, so a long title (or an
unbroken path/identifier) pushed the scope, date and action past the
container and made the Forget label break vertically. Let it shrink and
wrap within 60% of the row, and keep the trailing action on one line.
Withdraws the studio-cli-channels environment from #11395 at the owner's
decision: no human approval. The job reads the shared NPM_TOKEN and
HOMEBREW_TAP_TOKEN. The contract test and STUDIO_RELEASE.md follow.
cli-channels runs in the studio-cli-channels environment and is the only job
with id-token: write. Its registry tokens are environment-only secrets
(STUDIO_NPM_TOKEN, STUDIO_HOMEBREW_TAP_TOKEN), and its first step fails unless
the environment has a required reviewer. npm publish passes --provenance,
matching the 1.x release line. A contract test parses every workflow and pins
the permissions, the secret references and the id-token grant. npm/ changes now
trigger the CI job that runs npm/publish.test.mjs.
neterr.IsConnReset fell back to any net.Error, so a *net.DNSError read as a
dropped stream and users saw "disconnected before completion after retry
attempts" for a host that never resolved and was never retried. DNS errors are
now excluded from the reset judgement, and explainError gives them their own
typed branch that names the host and separates a nonexistent name from a
temporary resolver failure.
Refs #11391
The storage panel showed "Measuring..." forever when one root (a huge or
slow directory, common on Windows) never finished, because GET /storage
walked all five roots synchronously with no limit and the client had no
timeout.
- Survey walks each root under a 10s budget and reports a lower bound with
Truncated set when it runs out, even if a filesystem call never returns.
- GET /storage?layout=1 answers at once with unmeasured roots and
GET /storage?root=<id> measures one root, so a slow root no longer holds
back the others.
- The panel draws rows immediately, fills each root in as it arrives, marks
a truncated count as a lower bound and shows "timed out" for a root the
kernel never answers.
Port of #11386 (main-v2). The recovery block now ends with a sentence
stating that the user's new message takes precedence, so an unrelated
message no longer makes the model resume the interrupted task.
Refs #11381
The docked workbench hides its canvas while the file list is open, and the
browser page and its address bar are drawn in that canvas. Selecting a tab or
opening a new browser tab left the list open, so the browser never appeared.
Every selection now goes through one helper that also closes the list.
The authoring guard skips any task without a solution/, so all twenty
verification-stress graders have only ever been checked against the pristine
seed - never against an answer that should pass. Ten now carry one.
The reference-solution test holds graders to both halves of the authoring rule
but skips memorybench, which ships no solution/ - the half #11329 fell through.
`-policy` is the MemoryBench experiment arm: `memory-off` hides the memory
store so a run can be compared against itself, and the only switch that
separates the two arms of that counterfactual. It is defined at
`cmd/e2ebench/main.go:268` and used in the recipe further down this README, but
the flag table never listed it.
Placed with the other suite-mode experiment arms (`-anchor`, `-cache`) rather
than beside `-ablate`, so it does not collide with the open #11364 that edits
that row.
ablation.Modules() has seven modules; the sentence documenting --ablate stops at
upstream. #11360 edited the same sentence to drop full-fold but left this one.
The bar named a field that does not exist (`PostSufficientSearches`) and a mode
that cannot produce counts (`-mode=calibrate`). The gate is the run's
`stopping_class` from the `-mode=run-index` report: a task qualifies when it is
neither `PostSufficientRetrieval` nor `PostSufficientRunaway`, so `SnippetStop`,
`SnippetThenRead` and `DefensiveRead` all qualify, and a defensive read is
deliberately not counted as waste. Also records that calibrate is not a
screening tool, which is why its cue-visibility profile is the same for every
task.
CloudDesktop only set Content-Type when the request had a body, so a
bodyless POST such as /runtimes/{id}/close reached csrfGuard without a
JSON type and was refused with 415 request.bad_content_type. Deleting a
project or session closes its panel first, so the delete failed in the
web Studio. csrfGuard is unchanged.
The benchmarks README still described this harness as the pre-#8244
comprehension/TTL/prune A/B, and taught a subcommand and flags the code has not
had since that rewrite.
A full-screen repaint draws every settled block, so when a background
[cli].diff_formatter run landed and the re-render hook was keyless, the
repaint dropped every diff block's cached rows and re-consulted the memo
for each. Once the transcript held more diff keys than the memo, the repaint
always re-missed at least one block, spawned a run for it, and that run
landed and repainted again — an endless loop of formatter runs and full
repaints.
Carry the landed DiffKey on the hook and remember, per block, the keys its
last render consulted, so a landed run drops only the blocks that asked for
its key. Gate the drop on the memo still holding the key: re-rendering a
block whose key was evicted would re-run the formatter and evict another
live key, so an evicted block keeps the rows it already shows.
The memo is raised to 128 so a normal transcript does not reach the eviction
path, and keyed by a 128-bit hash of the diff text rather than the text
itself, so an entry retains only the formatted output (~33-41% less
retained heap). Also record the formatter's 1 MiB size cap as a fallback
reason in the CLI reference.
ablation.FullFold has had no reader since #8244 (397ed30ec) replaced the
multi-threshold path with the single checkpoint design: foldSource and
carryPriorDigests lost their Off(FullFold) checks together with their test, so
-arm=incremental was identical to the default, and the default already folds the
model-visible projection. The checkpoint design is the intended one, so retire
the dead axis rather than restore the old re-read.
- ablation: drop the FullFold module (constant, comment, module list entry)
- compaction bench: drop the -arm flag, the arms struct, foldArm, and the arm
label from the report title and the JSON report
- benchmarks/README.md: drop the "Fold arms" section, and correct the paragraph
that claimed a fold re-derives its digest from the canonical transcript
- docs/CLI.md: drop full-fold from the --ablate list
Verified: go build ./...; go vet; go test ./benchmarks/compaction/
./internal/contract/ablation/; go run ./tools/repolint clean; golangci-lint
v2.12.2 (the version CI pins) 0 issues. -mode=cost still reports the same eight
generations, -arm is now an unknown flag, and the JSON report no longer carries
an arm label.
Fixes#11358.
studio's config.go sits exactly at its file-size budget, so the diff-fences
feature's added [cli] fields tipped it over repolint's ratchet. Extract the
CLIConfig type into cli.go, alongside the existing render_cli.go, to keep the
budget.
A shell tool result whose whole output is a unified diff — bash running
`git diff`, or `git show` in its default format — is tagged and rendered as a
diff: the TUI draws it through the same rows a ```diff fence produces, ACP
fences it for its client, and the desktop card renders it with DiffView.
The tag is computed once, at the one sink every emitter shares, and is gated by
the opt-in [agent].embedded_diff_detection (default false). Detection is
whole-text and conservative (diff.IsUnifiedDiff): every line must belong to the
diff grammar — including a `git show` / `git log -p` commit header and its
indented message — and a lossy Bound is never tagged, so a diff followed by
prose, a build log, or a truncation note stays prose. The tag is re-derived when
a session is rebuilt from /history, so a reopened session renders the same diff
the live one did.
The desktop DiffView splits a multi-file diff into per-file sections by the same
rule the CLI uses (git's "diff --git", else the "--- "/"+++ " pair) and names
each with its own "path +A -B" header row, so a `git diff` over several files no
longer shows only its first file's name over rows that belong to all of them.
A single-file diff keeps the card headline as its only name, so the fact is not
said twice.
trimDiffPreamble returning "" for a section with no --- line left a ```diff
fence without a --- /+++ pair rendering as an empty block. splitDiffSections now
returns sections untrimmed; the fence path trims each and falls back to the plain
code rail when it carries no file header, while the shell-diff path still drops a
pure preamble. A marked shell diff keeps the card's "⎿" connector on its first
row, like every other tool card's body. /history reads the embedded-diff flag the
session was built with instead of reloading config per request, and the hunk
parser rejects a body that overshoots its declared count.
[cli].diff_formatter names an optional external argv line (run without a shell)
that formats a diff before the CLI/TUI emits it — both a fenced
```diff/```patch block and a writer tool's diff card. The whole diff goes to the
command's stdin and its stdout is re-emitted with non-SGR control sequences
stripped. On failure, timeout, empty output, or an oversized block the built-in
renderer is used. Like [cli].update_channel it is user/global only: a
project-local reasonix.toml cannot run a command on the user's machine.
The run is memoised per (block content, width), because a View re-renders many
times while a fence streams and again on every resize. It runs only on a closed
fence, so a fence still streaming spawns no subprocess per delta, and the command
is handed the terminal width via COLUMNS with its output clamped to it. A
WaitDelay bounds the wait for a formatter's pipes, so one that leaves a child
holding stdout cannot block the render past the timeout. The writer card drops
the "--- /+++ " pair before piping, so the formatter cannot re-emit a raw host
path the card header already names.
This commit also wires [cli].diff_fences, the opt-in toggle that turns the
fenced-diff renderer on.
A fenced ```diff or ```patch block in the answer stream is drawn through the
colourised diff renderer — add/remove backgrounds, +/- gutter, line numbers,
chroma — instead of the generic code rail. A multi-file git-style fence is split
into per-file sections, each with a bold(accent(path)) +A -B header. Copy mode
keeps the plain rail path so the copied text stays byte-identical to the fence
source.
A fence that already carries ANSI (a pasted colourised diff) is shown verbatim,
with non-SGR control sequences stripped so a hostile payload cannot drive the
terminal.
It is off by default: the renderer runs only when activeDiffFences is set, so a
fence a model writes stays on the plain code rail unless the user opts in.
`@scope/[email protected]` and `[email protected]` were rejected by both
looksLikePackage implementations (Go installer and registry publish
validation), so an npm MCP entry could not pin its version. Both now accept
an optional `@<exact semver>` suffix and still refuse ranges, dist-tags,
empty versions, a repeated `@`, and anything with whitespace, shell
metacharacters, path segments or scheme prefixes.
A leading `-` is now refused as well, so a bare source can no longer be read
by npx as an option. The default MCP name drops the version suffix.
Pinning the top-level package does not pin its transitive dependencies.
Add owner headers to twelve contract and design documents, split their
over-wide paragraphs into rules, route their ownership to the primary
named in CODEOWNERS, and treat the generated protocol index as a
machine-read file. Baseline: doc-owner 40 -> 27, doc-prose 2276 -> 2147.
Add idle_timeout_seconds, a per-provider override of the 300s stream idle
watchdog. One knob bounds both windows a model call can spend silent: the
transport's pre-header deadline (prefill) and the read loop's inter-event
watchdog (generation). Zero or unset keeps the default.
The bound holds where the value is read: IdleTimeoutFromExtra clamps to
[1s, 32767s], so a hand-edited value — which never passes UpsertProvider's
validation — cannot overflow time.Duration(v) * time.Second into a negative
window the transport reads as "no timeout". validateProvider still rejects an
out-of-range value on the edit path.
The stall error names idle_timeout_seconds, and the docs note that a silent
pre-header wait is retried once, so the effective hold can reach ~2x the value.
A Studio install on another drive than the update cache never took the
chunked update: the swap's pre-flight renamed a probe from the install
into the cache, and a rename cannot cross volumes. The swap backup now
lives inside the install (<install>\.reasonix-update\<version>), the one
place guaranteed to share its volume; staged files still come from the
cache and are copied in when the rename fails.
Chunks are fetched through update.Transport, so they take the IPv4
route from the second attempt as the full download does, and chunks a
round could not fetch get a second round instead of voiding the delta.
Why a delta was abandoned is now a typed code (update.delta.*) that the
step giving up attaches; it rides on Progress.DeltaSkipped through the
full download and its failure, is logged, and the version panel says it
is downloading the full package and why.
Full downloads drop a connection that delivers nothing for 30 seconds
and resume on a new one; only attempts that made no progress count
against the retry budget. The artifact is written to a partial file in
the update cache named by its digest, so a timed-out or restarted update
resumes instead of starting over; a partial that fails verification is
removed. A body the manifest's size rules out (an error page served as
200) fails that address and moves to the fallback without being resumed.
With stalls caught, the fetch deadline is two hours rather than thirty
minutes.
A file copied in from another volume is written under a temporary name
beside its destination and renamed into place, so a crash mid-copy never
leaves a half-written file at the install path. A backup directory the
install refuses to create is ErrTreeNotSwappable. The delta phase has
its own 40-minute budget (update.delta.timed_out) and gives up after 16
chunks fail in a row, so a mirror that has gone quiet leaves the full
package its time. The reason is also exposed to screen readers.
"pnpm install" contains "npm install" as a substring, so verify.sh's
negative grep matched the very answer the positive grep demanded, and the
task could never pass. Disambiguate the negative check so it only matches
"npm install" when it is not preceded by "p", instead of anchoring the
positive check to line start, which would have moved the grader out of
line with every sibling task's plain substring convention.
Also extends TestSolvableCorpusSeedsMustNotGradeClean to the memorybench
corpus (it was not covered by the existing e2e/verification-stress/train
authoring guard) and adds a targeted regression test pinning the four
cases worked through by hand in the report.
Credit to chuiwenwei for the root-cause analysis and both candidate fixes.
Fixes#11329
The Studio manifest named only GitHub release URLs for full artifacts, so
every update downloaded the installer over the route CN users reach
slowest, even though the release workflow already uploads the same files
to the R2 mirror.
When the release is mirrored, studio-manifest now writes the mirror
address as each asset's url/sig and keeps the GitHub address in new
optional fallback/fallback_sig fields. Old clients ignore the new fields
and use the mirror, which the redirect allow-list already accepts.
The updater tries each address in order with its own retry budget,
resuming what the first delivered from the second. Signature, digest and
size checks are unchanged and apply to whichever address served the
bytes.
An approved market package whose reviewed version carries no content digest
was a dead end: the detail page showed "not pinned" and no way to install it.
It now offers "Trust and install". The request carries trust, and the kernel
plans it through install_source exactly like any other market install; the
preview's own contentDigest becomes the pin, apply must echo it, and content
that changes after the person confirmed is refused. The outcome and the ledger
record the install as unreviewed, and the confirmation says the content may
differ from what was reviewed.
A pinned version is unchanged: trust never replaces the reviewer's digest, a
digest other than the pin is refused as a version change, and moved content
still fails with market.content_changed. Without trust an unpinned version
still answers market.unpinned.
The publisher's own unreviewed install and the trusted install now share one
preview-pinned path, and market.unpreviewed is answered on both routes. The
list filter and badge read "pinned" instead of "installable", since an
unpinned package is now installable too.
A command the host cannot classify gets a before/after workspace scan,
and a scan that sees a change proves a write. An inline plan check such
as `python -c "import fizzbuzz; ..."` imports the module and writes its
bytecode, so the check was proven a change and owed itself another run.
Each check now measures "after the latest change" from the latest proven
change not counting its own residue: a run of exactly that check whose
written paths a complete workspace walk named, each of them a file whose
first touch this turn was its creation, none of them a file the plan
names or any other write this turn touched. A check that deletes a file
on one run and writes it back on the next is not its creator. Anything else counts: a run that does more than the check,
a check that rewrites the source it just checked, a check that rewrites
any file that predates the turn whether or not anything named it (go
generate, a whole-tree formatter, a lockfile rewrite), a path list cut
at the observation limit or never walked, and every run of
a different check - so a rebuild by one check voids another check's pass
taken against the stale build, and a check that never ran is not settled
because the only write came from another check. The plan deliverable and
the change criterion leave out the residue of every plan check.
The walk now also runs after a call whose watched paths already proved
a change, and records which files it saw appear: a new top-level entry
proves a change on its own, and without the walk a rewrite deeper in the
tree next to it went unnamed. The receipt says when its paths came from a
complete walk (paths_complete), so residue is never read off a list that
only holds what the host happened to watch.
The executor handoff now says what counts for a verify line: that exact
command, run as written after the latest change. It used to invite the
executor to validate and adapt inferred commands, which is how most held
runs ended up running a reworded variant.
Move the ProviderEntry/ProviderModelOverride types, their accessors and
validation into provider_entry.go, and the provider table rendering into
render_providers.go. Trim the moved doc comments to the <=3-line budget and
reclaim the freed repolint budget.
No behavior change: same types, accessors, validation and rendered output.
The TUI laid rows out with grapheme widths while bubbletea's renderer counts
per rune until the terminal answers the mode 2027 query, and so do the
terminals that never answer it (Terminal.app, legacy conhost). A VS16 emoji
pulled the scrollbar left, a ZWJ row lost its scrollbar cell, table lines
moved, and on conhost or behind an approval card those rows overflowed and
every row below was drawn twice. conhost also drew emoji as fallback glyphs
and `◆ ●` as `?`, and Windows Terminal drew boxes for flags and keycaps.
1.x has the alignment defect as well.
- termrender measures through one method, Cells(), with per-rune Truncate
and Cut of its own (the library's cut whole clusters and came out wider
than asked). The TUI sets it per rune and switches it to grapheme clusters
on the mode report bubbletea acts on; before the program starts it asks the
terminal how far U+26A0 U+FE0F advances, so one drawing it two wide gets
the grapheme count without answering 2027.
- Per rune, the whole frame is drawn stand-ins that keep the per-rune count:
joiners, selectors, keycap marks and tags go, a skin tone becomes its two
blank cells, a regional indicator its letter, and a one-column emoji with
only a two-column glyph is kept only before a blank.
- On conhost, told from a pseudoconsole by its window class, each rune is
asked of the console's own font (glyph and advance) and measurement; one it
cannot draw as counted takes a drawn symbol of the same meaning, its
halfwidth form, a flag letter, the OS best-fit when it really maps, or
blank cells. Windows Terminal gets letters for flags and base characters
for keycaps and tag flags.
- Live rows full screen are drawn at the transcript's width, so a pending
tool row is not wrapped again when it lands there.
Selection copies the original text; stand-ins change only what is drawn.
A desktop session whose person never named a posture now opens the way the
terminal does since #11256: Auto only where the OS sandbox confines writes
and the folder is trusted, Ask everywhere else (always on Windows). A named
posture - set in config or switched on the composer - is kept, and a value
this build does not know still opens in Ask with its notice.
- control: mode and "defaulted" are one approval sub-state; a trust answer
re-derives a defaulted posture with one check-and-set under the approval
lock, so a concurrent switch stays named. New FolderTrust port (Posture,
DecideWorkspaceTrust). TrustableFolder resolves symlinks and refuses a
filesystem root, anything holding a home directory, and anything inside or
holding the Reasonix home or state directory.
- config: desktop.default_tool_approval_mode is unset by default and renders
as a comment. A startup step releases a written "auto" once, recorded in
desktop-posture-released.json beside the config rather than in
config_version, which the 1.x line's own upgrades own; ask, yolo and
unknown values stay. The studio host now runs the startup upgrades.
- serve: POST /workspace-trust, approvalDefault on /status; the hub carries
only a named posture to later panes; a rebuild re-derives a default.
- studio: the terminal's "trust this folder?" question beside the composer
where a sandbox would confine what trust allows, and a note with a
one-click switch to Auto where no sandbox is in effect.
1.x's `reasonix run` and `-p` leave stderr empty on a clean run. Studio wrote
three slog records there on every run: `INFO boot: assembly timing`,
`WARN prompt refiner provider construction failed` (whenever the optional
refiner side model cannot be built) and, with -c, `INFO controller: resume
cache state`. A script that treats stderr output as a failure, or captures it
into a report, broke on every invocation.
All three are diagnostics nobody has to act on, so they log at debug level.
`run --debug` lowers the process log level while flags are parsed, before
assembly writes its first record, for whoever wants them back. Warnings the
user has to act on keep their own channel.
/events numbers frames from where a subscriber attaches, and Subscribe
connected in the background. An action taken right after it returned could
emit its frames before the connection registered, and the stream never
carried them. TestUnknownSlashCommandNeverReachesTheModel lost its
unknown-command notice this way whenever the kernel answered faster than the
subscription attached (every time on a Linux runner with GOMAXPROCS at 1 or 2, and on the coverage job), and the TUI's
own Init has the same order: it subscribes and then submits --prompt.
Subscribe now returns once the first connection has a response in hand (the
server registers the subscriber before it writes a byte) or has failed.
Every user-config save (reasonix config compact-ratio, the Studio settings
page, the serve settings endpoints, model and approval persistence) went
through Config.SaveTo, which re-rendered the whole file from the struct.
Anything the struct does not decode was dropped - [checkpoints],
browser.chrome_path, providers[].display_name, agent.web_search_model,
desktop.layout_style / status_bar_style_initialized,
remote.hosts[].credential_mode - and values the renderer normalises were
rewritten, so a 1.x default_tool_approval_mode = "danger-full-access"
became "ask". Losing credential_mode = "remote" flips a remote host from
using the host's credentials to tunnelling the local key.
An existing user config is now edited in place: the file is indexed by
table path (array elements by their name), the rendered form of the file
as it loads is compared with the rendered form of the config being saved,
and only the keys that differ are rewritten, added or removed. A key the
save changed but the renderer never writes (a retired field a setter
cleared) is removed, and the keys this build declares retired are still
stripped. Unknown keys and tables, comments, order, line endings and
values this build reads differently stay byte-for-byte. The edited text is
accepted only if it loads to the same configuration the full render would;
otherwise the full render is written and the previous bytes are kept
beside it as config.toml.rewrite-<time>. A missing or empty file is still
written from the template.
A config written from nothing now carries config_version = 12. The 1.x
line rewrites the whole file on start when the marker is below its latest
upgrade (12), dropping [storage], [memory] and the other tables it does
not know. This build's own upgrades still gate on version 6 and an
upgraded older file is still marked 6, so 1.x keeps running its 7-12
migrations for content that never had them.
Arrays of tables are edited per element, keyed by name. The whole array
is written from the render instead - leaving every other byte alone -
when an element the save changes or drops is not a block of its own in
the text, when an element has no name or shares one, when the save
reorders the elements, or when the file holds no such array yet (so
adding the first provider keeps the built-in ones). A new element is
placed before the next kept one, so a rename keeps its neighbours whole.
Dropped elements are removed in one pass over one index, so an element's
sub-tables are never handed to the element above it and removing many
elements is linear.
Keys and sub-tables Config does not decode are never dropped silently.
Rewriting an array carries each element's own into the rendered element
of the same name, verbatim; turning an inline table into a table carries
its plain entries as key lines. When an entry has nowhere to go (two
elements share a name, an element the load merged away, a nested table
value) the result is still written in place and the prior bytes are kept
beside it as config.toml.rewrite-<time>, as the full-rewrite fallback
does. An element the user deleted takes its entries with it.
A value the user sets is written even where the render would not show a
difference: a single-line key is rewritten when its encoded value
changed, so choosing Ask replaces a 1.x danger-full-access that already
loads as ask. A key nobody touched keeps its bytes.
A save never writes a file that does not load. system_prompt is
rendered as an escaped multi-line string (a backslash or """ in a prompt
used to produce an unparseable file). When the render still does not
load, the in-place edit is used if the lines the decoder refuses are the
same before and after the edit and the result loads as the render
describes; otherwise the save fails and the file is left as it was.
Inserted text takes the file's line ending outside strings only.
An unrecognised default_tool_approval_mode still opens a desktop window
in Ask, now with an approval_mode_unrecognized notice the frontend
localizes, and the file keeps the value until the user picks a mode.
The fixture is the file 1.x v1.39.5 writes through its own SaveTo with
the affected keys set.
A one-word message let the edit and rewind controls widen the card, and the
bubble then sat against the card's left edge, leaving a gap on the right.
The controls row no longer sizes the card (contain: inline-size), the bubble
is pushed to the right edge, and the controls drop to icons with their
accessible names, hints and keyboard focus when their labels need more room
than the bubble gives. Whether they fit is measured, not a fixed breakpoint.
Closes#11255
In the reading measure the transcript was 760px wide but replies and user
turns were still capped at --doc (706px), a value left over from when the
flow was a wide container with a narrower prose measure inside it. The
steps card inside a reply and the user bubble therefore stopped 60px short
of the composer's right edge, while the ask card and notices reached it.
The column now has one source: --column (760px) in tokens.css, with --doc
derived from it. The flow is laid at --doc, the composer and its shelves
at --column, and the per-row caps are gone, so every card shares the
composer's edges. The full measure still only moves --doc: the transcript
widens and the composer keeps the reading width. The studio-only
--measure variable that duplicated --doc is removed.
cards.mjs asserts the shared edges at 1440 and 1000px in the reading
measure, and that the full measure leaves the composer unchanged; the
new checks fail against the previous stylesheet.
Closes#11267
In 1.x a ! command is the user's own: it runs, its output is shown, and
nothing enters the conversation. The TUI handed it to RunShell, whose
answerShell puts the output into the conversation and starts a turn, so every
command cost a request and the model often replied with a question.
RunShellWith takes a ShellRun; LocalOnly skips answerShell. It sits on a new
LocalShell sub-port that only SessionAPI carries. serve's in-process /submit
takes localShell, and the TUI's shell mode sends through a RunShell client
call that sets it and echoes the command without waiting for a turn to name
it. RunShell keeps answering, so Studio is unchanged.
RunShell and its helpers move from controller.go to shell.go, which the
file-size ceiling on controller.go required.
With no terminal, 1.x printed usage for a bare `reasonix` and exited 0;
studio exited 2, the usage-error status, although nothing was wrong with the
command line. Session options with no terminal to open a session in remain a
usage error. bareUsage moves out of cli.go into its own file.
`reasonix doctor` warned on every install that the built-in review and
security-review skills allow use_capability, "which is not in the current
registry". Doctor checked allowed-tools against a hard-coded name list and
passed no tools; assembly always registers the proxy. It now checks against
the compile-time built-ins and the proxy, read from their owners.
1.x failed a request to an unreachable model service in about 0.3s. Studio
counted every transport error except cancellation as transient, so a refused
connect walked the whole MaxRetries backoff and `reasonix run` took about
92 seconds to report it, stalling every automated check pointed at a stopped
endpoint.
A refused connection (ECONNREFUSED, or WSAECONNREFUSED on Windows), a name
that does not resolve, and a certificate that does not verify are now
permanent: no backoff changes them. They are recognised by error identity.
Resets, timeouts and retryable statuses keep their retries.
1.x's `reasonix run` exited 0 once the model finished. Studio exited 1 with
`error_during_execution` whenever its final-readiness check stayed unmet (for
example, no check ran after the last write), so every 1.x script treated a
finished run as a failure.
Following the migration's decision 3 (in line with how other agent CLIs
report this):
- An unmet final-readiness judgement is a verdict, not a failed run: subtype
success, is_error false, exit 0, with `readiness` {attempts, missing} in the
result and in events-jsonl's run_done.
- `--fail-on-unverified` turns that verdict into exit status 3, kept apart
from 1 (run error) and 2 (usage) and documented.
- `permission_denials` in the result stays the gate's own list; run_done
carries its count and `-p` names the refused calls on stderr.
- `completion` repeats the turn's completion_summary when there was one.
1.x's `run --output-format stream-json` stamped every line with sessionId,
turnId, seq and status, opened the turn with turn_status (queued) and
user_message, reported tool_started between a call's dispatch and its result,
and closed with turn_done; --events-jsonl carried the same lifecycle records
without content. Studio's synchronous controller path publishes no turn
lifecycle, so both streams lacked all four kinds and the envelope, and they
leaked host-internal kinds (workspace_lease and friends) that no 1.x consumer
knows.
The run's output sink now supplies the lifecycle for the turn it drives and
stamps the envelope; only the kinds 1.x emitted reach either stream.
tool_started comes from a new host-only Tool.Executed on ToolResult, set when
the call cleared every gate and ran, so a refused call reports none.
The stats recorder swallowed RecordTurnCompletion instead of passing it to the
sink it wraps, so the run sink never learned a turn had finished: a failed run
reported num_turns 0. It now forwards the signal once (a TurnDone event still
records without a second completion).
Contract tests compare a real run against v1.39.5 output kept under
testdata/v1-run: lifecycle order, envelope keys, kind vocabulary.
* fix(goal): align Studio schema with report validation
Problem: schema-valid goal reports can fail execution for missing reasons or
blank verification commands.
Root cause: status-dependent requirements exist only in prose and execution
checks, not in the provider-visible JSON Schema.
Fix: encode conditional reason requirements and nonblank verification entries,
matching the executor's Unicode whitespace handling without relaxing Goal
authority or delivery validation. Update bilingual guidance and schema fixtures.
Verification: schema/executor agreement regressions and golden guards pass,
along with go vet ./..., make lint, and make check. The static schema changes
once on upgrade; system prompt bytes and within-version stability are preserved.
* fix(goal): keep report schema typed and within prefix budgets
The conditional status schemas omitted explicit types, which failed the
built-in provider compatibility gate. The added constraints also exceeded
both schema token budgets by 78 tokens.
Declare string enum and object branch types and trim descriptions whose
semantics are already explained by the tool. Keep every argument constraint
and the original budget ceilings, and refresh the serialized golden files.
Verified update-goal schema/execution tests, enum/type and schema-budget
guards, provider golden guards, go vet ./..., make lint, and make check.
* test(goal): keep whitespace validation coverage without schema changes
Problem: the schema hardening introduced provider compatibility risk through
a top-level anyOf and a new pattern constraint, and reduced model guidance
to fit the prefix budget.
Root cause: matching a general-purpose JSON Schema validator to execution
does not establish acceptance by model providers.
Fix: restore the original schema, descriptions, documentation, and golden
fixtures. Retain focused executor tests for blank reasons and verification
entries, optional completion fields, field diagnostics, and no report on
rejection. Existing error messages already identify the invalid fields.
Verification: update-goal tests, go vet ./..., make lint, and make check pass.
The owning suites retain the baseline ripgrep failure addressed by #11244.
The Regenerate and More menus on a finished reply were absolutely
positioned above the trigger inside the card. The card paints under
content-visibility containment and sits in the transcript scroller, so a
row near the scroller's top lost the upper part of the menu, and a short
reply lost it even mid-screen.
Both menus now go through one ReplyMenu: the list is portaled to the body,
placed by measurement, opens above the trigger when it fits and below it
otherwise, and follows the trigger on scroll and resize. The row keeps a
focus stop where the list used to sit, so Tab order through the menu is
unchanged; Escape now also returns focus to the trigger.
The scroll/resize follow effect that Picker and RewindControl each
carried is now one useFollow hook in place.ts, beside the shared beside()
flip arithmetic, which lowers the unnamed-input ceiling by one.
Closes#11250
/help was not a command, so it went to the model. /compact drew one line,
"context compacted", even for a fold that wrote nothing, and /context
dropped the breakdown the kernel attached to its summary.
/help is now answered on screen in 1.x's layout (commands, then built-in,
custom, skills and MCP prompts, eight rows a dynamic group), built from the
completion catalogue the session answers, the TUI's own commands included.
A finished fold draws 1.x's card: messages folded and trigger, the digest,
and the estimated tokens before and after (i18n CompactionEstimatedTokens,
the 1.x key); a fold with no digest draws none, and a running one says it is
working. The kernel names /context's notice with a context_report code, and
the TUI shows its detail under it.
Problem: A hidden directory could make grep search HOME and attribute its
matches to the requested external directory under Linux bubblewrap.
Root cause: Searching dot relied on the wrapper preserving the subprocess
working directory, but bubblewrap can fall back when that directory is masked.
Fix: Resolve directory roots to absolute physical paths and use the same path
as both cwd and the ripgrep argument. Restore requested output prefixes and
external-root error displays while preserving wrapped error identity.
Verification: Real Debian bubblewrap reproduced the prior failure for both
tmpfs and session-temp mounts; both cases pass with the fix. Grep regressions,
builtin and boot package tests, go vet, make lint, and make check pass.
The prologue was one chained assignment,
`$OutputEncoding=[Console]::OutputEncoding=[System.Text.Encoding]::UTF8`.
When the Console assignment throws (no console attached, or
ConstrainedLanguage refusing .NET property writes) neither encoding was
set and every command printed a 7-8 line error first. Each assignment is
now its own try/catch and .NET properties are only touched in
FullLanguage. Both encodings are BOM-less UTF8Encoding instances, so text
piped to native programs no longer starts with EF BB BF.
On Windows PowerShell 5.1 only, the shell tool also sets
$PSDefaultParameterValues['Out-File:Encoding']='utf8', so '>' and Out-File
write UTF-8 instead of UTF-16LE. '*:Encoding' is deliberately not set: it
would make Get-Content misread ANSI (GBK) files. Hooks keep only the output
part of the prologue, since they run the user's own scripts.
PowerShell discovery adds the Store install's execution alias
(%LOCALAPPDATA%\Microsoft\WindowsApps\pwsh.exe). An alias is a zero-byte
reparse point that stats fine but may fail to start the packaged app, so an
alias candidate (from the list or from PATH) must launch once before it is
selected; regular executables are taken as found.
A Chinese terminal still read English in the rows the TUI or the kernel wrote
themselves: "remembered → …" after a # note (1.x says 已记住), the session
picker's Search / No matches / Type to filter, a declined approval, a
sub-agent's call count, and the progress-watch notices. The completion menu
showed its key hint only past eight rows, and in English; 1.x always shows
it, from CompHintSlash / CompHintFile.
These now read from i18n, reusing 1.x's keys where 1.x had them. Text 1.x
itself prints in English (N lines, (active), N turns) stays English. The
language detection moves from i18n.go to detect.go; i18n.go was at the
800-line ceiling and the new keys need the room.
The memory-off arm hid the memory tool but still seeded the store, and the
state root was a sibling of the task workdir under the temp root, so agents
in the off arm found and read the seeded facts with shell commands.
- memory-off no longer seeds memory: the treatment is that no memory exists
- state roots live under the user cache dir with a random name, and a root
inside the workdir's parent is refused
- -trajectories is made absolute before the child, whose cwd is the task
workdir, receives it
A bare /rewind went to the kernel, which rewound the conversation one turn at
once, and /clear deleted the transcript with no question and left it on
screen. Both are destructive; 1.x asks first. Esc on an idle composer did
nothing, where 1.x clears what is typed and a second Esc opens the rewind
picker.
The TUI now answers /rewind with 1.x's picker (turn, then code+conversation,
conversation, code, or a fold before/after the turn; a partly covered file
restore asks again), prefills the turn's prompt and redraws the conversation
after a conversation rewind. /clear opens 1.x's Clear/Cancel question, the
cursor on Cancel, and clears at once in YOLO; the screen restarts under the
banner. Esc clears the composer, and twice within 600ms on an empty one opens
the picker. The kernel's /new and /clear notices now come from i18n.
Every todo_write and complete_step call drew a row ("● todo_write ⎿ 1
lines", "● Step"), though the task list above the composer already shows
what they changed; 1.x draws neither. A shell row previewed the result the
model reads, and the host appends its obligation notes to that result, so
"host obligations changed: - settled: …" appeared as if the command had
printed it.
A settled todo_write or complete_step that succeeded is now kept off the
screen, as 1.x keeps todo_write; a failed one still shows. A shell row
previews what the command streamed while it ran and falls back to the result
only when nothing streamed (a replayed session).
Found in an adversarial review of the new default posture:
- shellsafe: arguments that make a read-only command run another program or
write are refused in every spelling the program accepts (clustered short
flags, abbreviated long options): git grep -O, rg --pre, sort
--compress-program, go -toolexec/-exec/-vettool, file -C, uniq's output
operand, date -s, hostname NAME, git reflog expire/delete, npm audit fix,
kubectl --kubeconfig, gofmt -cpuprofile. less, man and cargo check/doc
leave the read-only tables.
- The headless gate and the posture decision live in the files REASONIX.md
declares sensitive again (approval_bridge.go, approval_orchestration.go);
the MCP fast-path refusal is decided in package permission.
- A sub-agent under an attended Auto no longer gets dynamic shell its parent
would have to ask about.
- Read-only keeps fresh human decisions human; a refusal no rule produced
names the posture instead of a deny list.
- The trust prompt and docs state what the sandbox actually bounds (temp,
toolchain caches, network) and that trust covers reasonix run; a home
directory or filesystem root is never trusted wholesale; the run result
names the posture it settled on.
- The desktop names each controller session in ready; a controller that
uses the name numbers its commands upward, and the desktop refuses a
command naming another session or an old number, so a relay cannot
replay a recorded command into a new session or twice into one.
- ready carries a per-process instance id. A state-changing request
already sent is resent only to that same instance; a restarted desktop
has no cached answer, so the request fails explicitly instead.
- The replay cache outlives controller sessions (the grace expiry no
longer clears it) and is reset only on device identity or account
change; its key binds method, path and body as well as the id.
- The phone seals and sends commands in one queue so their numbers
arrive in order; socket listeners are registered by the connection
itself; errors reach the banner through reason().
A terminal session with no named mode opens in auto only where the OS
sandbox demonstrably confines shell writes (sandbox.IntegrityEnforced) and
the person trusted the folder; everywhere else it asks, and a headless run
refuses writes and lists them as permission_denials in its result.
read-only is a real posture again: every call that is not a read is refused,
allow rules included, and an authorized MCP server's writers too. Gate
refusals carry a code (permission.unattended, read_only, deny_rule,
declined) that reaches the tool result and the run result.
YOLO keeps skipping prompts only; its first interactive use is confirmed
once and recorded under the Reasonix home. Shift+Tab follows 1.x's order,
Ctrl+Y returns to the posture it left, and a terminal's switch no longer
becomes the desktop's default. git -C/--no-pager classify as the
subcommand they wrap.
The kernel's slash menu drops the row that would replace the line with
itself, so a finished /tree left only fuzzy neighbours in the menu and Enter
accepted the first of them: /tree ran as /security-review. A finished /hooks
or /mcp stopped at their argument menu and a second Enter put its first
argument in the composer.
Complete now names the dropped row as Typed, which Studio ignores. The TUI
lists it first, and Enter sends the line when it already reads what the
highlighted row would make it: the command's label, or the argument in place.
That is the rule 1.x's composer follows; a half-typed name still completes.
Problem: Studio's ripgrep engine misses slash-glob matches under search
roots with symlinked ancestors, including macOS temporary directories.
Root cause: the absolute search argument keeps the symlink spelling while
ripgrep anchors globs against its physical process directory.
Fix: search dot from the requested subprocess directory and restore the
requested output prefix before external-root token projection. Both search
passes share the behavior; single-file targets and sandbox guards remain.
Verification: existing and new slash-glob regressions fail before the fix
and pass after it. Owning builtin/boot suites, go vet ./..., make lint, and
make check pass. The new test covers both passes and output path privacy.
The phone page reconnects silently when the relay closes for idleness, a
legacy admission expiry or a dropped socket, and resends in-flight requests
under the same id. A read is always safe to resend; a state change is
resent only to a desktop that advertises replay, and failed otherwise, so
nothing runs twice. Relay close codes decide the next step: 4401 asks the
user to sign in again (sign out, then the existing sign-in page), 4403 and
4004 stop for good.
The desktop keeps controller sessions across its own relay reconnects,
prunes them against the relay's relay_hello list, sends the relay's
heartbeat frame once the relay names it, and caches answers to
state-changing desktop requests for ten minutes so a resend is answered
from the cache instead of running again.
Refs #11226
The terminal typed /status, /help, /quit and the other 1.x commands studio
has not ported yet, and the kernel sent each line to the model as prose. That
spent a request, and for /status it ran git and asked for an approval.
A submission can now ask the controller to refuse a slash command no registry
resolves: it gets an unknown-command notice (i18n SlashUnknown, the wording
1.x uses) with the unknown_command code and starts no turn. Paths, code
comments and custom commands resolve exactly as before. The TUI always asks;
Studio does not, so its behaviour is unchanged.
A 1.x command line that reached one of these exited 2 on studio:
- `reasonix chat` / `reasonix code` were unknown commands. They are the
interactive session in 1.x, so they now route to the terminal UI.
- `reasonix --preset X -p task` was an unknown command because --preset was
not a leading session flag. It is again, so it becomes a print-mode run.
- `run --takeover` was an unknown flag. It is accepted; with no contention the
run proceeds as in 1.x. Taking a session away from another process has no
counterpart here, so a held session is still refused, and the refusal now
says --takeover cannot do it and points at --copy.
- `sessions reindex|diagnose|cleanup` fell through to the machine `session`
command. Sessions are listed straight from their directories here, so
reindex has no catalog to rebuild and reports the count it reads; diagnose
and cleanup report and trash covered conflict-recovery branches through the
sessionstore guards Desktop already uses, keeping 1.x's JSON key names for
what is measured and its dry-run-until---apply contract.
A provider entry gains display_name, the key 1.x already writes, so a
config shared with 1.x keeps it on every save. Name stays the identity:
refs, default_model, saved sessions and preset upgrades still point at it.
POST /providers/display-name relabels every entry shown as one account,
all or none, and refuses an over-long label or a control character with
its own code. The provider list, the model list and /status carry the
label; the settings list renames in place (pencil or F2, empty restores
the derived name), and the model picker, wallet header and ACP model
descriptions show it.
Closes#11237
resolveSetupTargets takes every unrecognized argument as the config path, so
`reasonix setup --help` wrote a file named --help and reported "Wrote --help".
1.x prints the setup usage for -h/--help and exits 0, and refuses any other
dashed argument with usage on stderr and exit 2. The ACP registry advertises
`reasonix setup` as the login method, so its help has to answer.
Port that guard, in its own file, ahead of target resolution.
The in-app find already searched the rows rather than the document, so
it sees messages the transcript has not mounted. What stood between a
person and it:
- The chord matched e.key exactly. Caps Lock sends "F", an input method
answers "Process", a Cyrillic layout sends "а" - none of them opened
anything. The window now reads the letter case-insensitively and falls
back to the key's position when the layout reports no Latin letter.
It takes Cmd on macOS and Ctrl elsewhere, so Ctrl+F in a macOS text
field keeps its Emacs meaning, and AltGr is not read as the modifier.
- Nothing on screen offered it. The title bar gets a find button (the
only way in on a phone) and the palette gets the command, both showing
the chord.
- It took the chord from every text field. A field hands the window's
chords through only when it says so (data-window-keys): the composer
and the find box do, a settings filter or the browser's address bar
does not. An editor that answered the press itself (CodeMirror's own
find) still spends it.
Stepping also moves within a row now. Twenty matches in one tool output
counted as twenty but painted and landed as one; the current occurrence
is the nth drawn in its row, and landing scrolls it - through the
output's own clip if it scrolls - to where it is read instead of leaving
the row's top under the bar. A newer landing cancels one still settling,
which typing a query used to race. Arrow keys step like Enter.
The current match was accent at 45%, which in the Studio theme is a
light grey behind light text in dark mode; it is the accent/accent-fg
pair now, neutral in both themes.
The window's key listener moved out of App.tsx into windowkeys.ts to
keep App under the line ceiling.
A workspace path concatenated into the sidecar is not valid JSON on
Windows, where its backslashes read as escapes; the meta then fails to
parse, the session is never routed to the blocked directory, and the copy
the test expects to fail succeeds.
A shell command whose command line exceeds CreateProcessW's 32,767 UTF-16
unit ceiling never starts; the model was told "fork/exec powershell.exe:
The filename or extension is too long." and read that as a file-size
problem. On Windows the PowerShell tool passes the whole script as one
-Command argument, so writing a large block of text (e.g. a Chinese
document in a here-string) hits this every time.
The host now counts the command line exactly as os/exec builds it and,
when it will not fit, refuses the call before launch with a typed error
(errCommandLineTooLong) that states the size, the limit, and that nothing
ran, and points at write_file/edit_file. The execution is recorded as
not run / not started. The not-run bookkeeping moves into one helper so
ExecuteDetailed stays within its size budget.
Refs #11229
Opening a session 1.x wrote deleted its <id>.context.json. 1.x writes
context schema 4 from 1.36.0; LoadCompactionState accepted only 1-3, and
loadProjectionSidecar removed the file on any load error. The model-written
compaction summary was lost for good, 1.x itself resent the full history on
its next turn, and 2.x sent the full history too.
A load error now drops the projection in memory only. Unknown schemas carry
ErrContextSchemaUnsupported. Schema 4 is schema 3 plus
projection.pinned_context_hash, so it is read when that hash is empty; a
pinned-context checkpoint is authenticated against 1.x provenance this build
does not model and stays unsupported. A load never rewrites another release's
file (lineage keys are normalised in memory), and dropping the projection of
an open 1.x event log leaves the file, since this build never changes that
log. When a 1.x conversation moves to a 2.x session, the projection is
inherited there in schema 3 if it still covers the transcript, so the first
2.x turn sends 1.x's summary instead of the folded history.
The effect test replays a schema 2 session captured from a real 1.38.7 build.
Problem: Clicking the Dock icon leaves a Studio window hidden after closing to tray.
Root cause: macOS activates the existing process without emitting second-instance, and the shell had no activate listener.
Fix: Route activation through the shared window reveal path and ignore reveal requests during shutdown. Extend native smoke coverage across activation, second-instance, and tray entry points.
Verification: macOS native smoke, Electron unit tests, frontend build, host build, go vet, and repository lint passed.
A phone reaching Studio through the account relay has no EventSource; the
page swaps in RemoteEventSource, which polls /events/replay over the
encrypted channel. It fell short of the stream it stands in for in three
ways:
- It found /replay by rewriting a URL that ends in /events. Since the
silence watchdog (#11216) a replacement stream carries ?lastEventId=, the
rewrite missed, and every poll asked for the live /events route, which the
desktop refuses over the relay. An idle pane went silent for 45s and then
stopped receiving events for good.
- It learned its starting position by asking for lastEventId=0, so each
attach downloaded the whole replay log (up to 8MB for a long running
session) through the relay and threw it away. Asking past the watermark
returns no frames and the watermark alone.
- It never said the frames SSE writes on its own, so a quiet stream looked
dead to the watchdog and a truncated log looked complete. A quiet poll now
restates the watermark, and complete=false becomes a stream_gap.
Separately, a reload always focused the first open pane, because which pane
the window showed lived only in React state. The relay's "reconnect" button
reloads the page, so the phone landed on the top project's first session
instead of the one it was on. The tab now keeps that id in sessionStorage
and takes it back when the pane is still open.
importJsonlSessions reported only artifact failures through hadArtifactFailure,
so a session whose own copy failed was skipped silently and the one-time import
marker still went down. The next launch then saw the marker and never retried,
leaving that session unimported for good.
Both copy paths now report the failure, so the marker stays off and the next
launch retries, matching what a failed artifact copy already did.
await_user's `need` is shown to the user as a one-line host notice. With
nothing bounding it, a model put the whole question there, numbered options
included, so the user got a wall of text to answer by typing instead of an
ask card they could click. The description now routes a pick among named
choices to ask, and the tool refuses a need longer than 200 characters with
an error that says to put the substance in the reply or call ask. Length is
a property of the argument, not a reading of its wording.
Closes#11220
The hand-back notice printed the model's `need` as one plain paragraph, so
its bold, lists and line breaks reached the user as literal markup. The
kernel now tags the notice with the await_user code and carries the need
verbatim as its detail; the text form stays for the TUI and CLI. Studio
titles the card from the code and renders the detail through the same
sanitized Markdown renderer replies use.
Closes#11219
Workspace configuration only takes effect once the workspace is approved,
so the perseveration boot effect tests reached the real provider instead
of the registered stub.
The perseveration guard fixtures are byte-exact recordings; checking them
out with CRLF on Windows shifts every offset the test asserts.
A phone paired over the LAN that sleeps, or whose Wi-Fi drops the link,
can leave its EventSource looking open on a socket nothing arrives on.
EventSource only reconnects on an error it is told about, so the pane
kept showing a live connection while no event reached it, and on plain
HTTP/1.1 each such dead stream also keeps holding one of the browser's
six connections to the host, starving the requests a session switch
makes.
The kernel already restates its position on every transport each 15s
(stream_watermark). The client now treats three silent intervals as a
dead stream: it closes it and opens a new one that resumes from the last
frame it folded (lastEventId), so nothing numbered is lost. Timers stop
while a phone sleeps, so becoming visible and coming back online check at
once instead of waiting for the next tick.
A workspace's reasonix.toml and .reasonix/settings.json are read as
untrusted input.
- [sandbox]: bash may only move to enforce and network only to false;
forbid_read and denied_domains add to the user's; allow_write entries
and workspace_root apply only when they resolve inside the workspace.
${VAR} in sandbox and proxy paths expands from the process environment
only.
- [permissions] mode, allow and allow_dynamic_bash, [network] proxy
settings and [desktop] default_tool_approval_mode are read from the
user config only; deny and ask add to the user's. An "always allow"
given in a workspace is stored under the Reasonix home
(project-grants.json) rather than in the checkout.
- Programs a workspace declares (hooks, language servers, providers,
tools.search.rg_path, tools.shell.path, [browser], tools.system_one
endpoints and the local Laya runtime) run only once approved for that
folder. An approval records the declaration and a digest of the files
it names, is kept under the Reasonix home, and is checked again before
each start; a change refuses the start with ErrProjectProgramChanged.
`reasonix trust` lists and approves them, and saving hooks from the
hooks page approves the block written. Model settings that named only
a held-back provider fall back to the user's choice.
- Single program paths may not point into the workspace, a temporary
directory, an allow_write root or a toolchain cache.
- Sandbox write roots and host write directories keep the identity they
had when first granted; a root whose identity changed is dropped with
sandbox.write_root_redirected or sandbox.write_root_changed. The
host-writable directory list lives in internal/base/hostwrite.
- Every ignored value is reported as a load warning with a typed reason
(Config.IgnoredProjectSettings); held-back programs raise the
project_programs_awaiting_approval notice.
The workspace explorer leaves every dot entry out, so folders such as
.reasonix, .github or a .gitignore could not be browsed from Studio at
all. The explorer head now has a toggle that asks the kernel for them.
GET /workspace/files takes hidden=1 to include dot entries in both the
folder listing and the search. VCS stores (fileutil.IsVCSStoreDir) and
dependency trees stay out either way. The default is unchanged: dot
entries hold tool state and local configuration, so they appear only for
a reader who asked, and the choice is remembered per machine.
The explorer's head moves into its own component so WorkbenchPanel
stays under the file-size ceiling.
Closing the last tab folded the docked workbench even when the file
explorer was open beside it, so dismissing a browser tab took the file
list away with it. An open explorer is now a reason to keep the column,
the same as any tab still open; the column still folds with its last tab
when the explorer is closed.
The speed meter already opens with a sparkline of recent throughput, so
the gauge glyph before it said the same thing twice and cost width in a
rail that is tight next to the composer. The cache and context meters
carry no leading icon either; the speed meter now reads the same way.
Under 760px the workbench collapses to one column and the stylesheet hid
the explorer unconditionally, so on a phone paired over the LAN (or any
window that narrow) the files button toggled to pressed over an empty
body and the workspace tree could not be reached at all.
At that width the explorer and the canvas now take turns in the single
column through the existing data-files toggle, the same way a docked
workbench already does. Picking a file still puts the list away and
shows the file.
The guard detects a degenerate generation loop — a model echoing one short block
of text or reasoning byte-for-byte — by default, and reports it through the
progress-watch channel as a new `perseveration` cause. Nothing is cut and nothing
reaches the model unless the caller opts in: a byte-identical block is sometimes
legitimate output (printing a value 300 times, a run of identical table or log
rows, a page of `=` or `A`), so cutting it by default was the wrong trade. The
notice is said once per loop episode — a scan that trips again while the previous
scan also tripped stays quiet, and the next scan that does not trip re-arms it.
The user's `progress_watch.pause` setting is what ends a run on a detected loop.
Cutting the stream and nudging an unattended retry is opt-in with
`perseveration_retries`, under [progress_watch] or per provider on a [[providers]]
entry (the provider value overrides the global default). Unset or 0 reports only;
a positive value also cuts and nudges up to that many retries; a negative value
disables the guard entirely.
The cut path records best-effort usage, lets extensions rule on the terminal
before it is persisted, and never executes the aborted stream's tool calls. It
collapses the loop to the phrase and its last repetition in what is shown and
stored alike, so the transcript follows the trim instead of silently diverging
from the stored turn — no history rewrite is needed. It then shows the user a
`[retrying (N) avoiding perseveration]` interjection, appends a host-authored
nudge through the mid-turn steer path, and retries. Once the budget is spent the
run stops with a typed perseveration pause and emits the `perseveration_loop`
notice, so a parent cannot mistake the loop for a clean result. The retry budget
lives on the run, so the next user message starts fresh.
Adds a boot-level effect test asserting what reaches the provider request and the
frontend sink, and updates the guide.
The version panel's install ran download, install and restart as one act:
once the signature checked, the application ended with no prompt. It also
wrote the pin before downloading anything, for every target, so an update
that failed (or a Windows chunked swap that rolled back after restart, the
2.20.x case behind #11115) left "pinned to 2.21.0" on a 2.20.0 install,
with its only way out disabled while any other move was in flight.
- StartInstall now stops at a verified, staged release (phase "ready").
POST /update/restart commits it; it is refused with
update.restart_busy while a local turn or background job is running,
unless the request says the person saw that (force).
- The pin is written when the install is committed, only for a rollback;
a move forward releases any pin. A failed or dismissed install puts the
previous pin back. A pin naming a build other than the running one is
released at launch.
- A failed move carries a dotted code from the step that failed
(catalog, download, signature, disk, no package, installer). A forward
move records itself; the next launch, still on the old build, reports
update.not_applied with the swap helper's reason, which the helper now
writes instead of printing to a stderr nobody reads.
- The panel explains each code, links the full installer where installing
once by hand is the way out, and drops wording that promised automatic
updates Studio does not perform.
Add an opt-in vi command mode to the composer prompt ([ui].commandmode =
"vi"). At the prompt Esc enters command mode for navigation and editing
rather than interrupting a turn; command mode owns the keys h l j k 0 $ ^
i a x I A p D and the mode switch.
The composer lives in internal/frontend/tui, so the mode is implemented
there against its textarea: vimode.go carries the command-mode keys and
the mode switch, input.go routes keys through it, and ask.go and
prompts.go keep Esc from dismissing an open question card or deciding an
approval card. Command mode consumes any key carrying printable text, so
a CJK/IME character is eaten rather than inserted.
Because the interrupt key now depends on the composer mode, the
running-turn hint can no longer name a fixed key: with ui.commandmode off
Esc stops a running turn, while in vi mode only ^C interrupts, since Esc
is a mode switch whether or not a turn runs. The thinking
status line therefore names the key that actually interrupts — "(Esc
cancels)" or "(Ctrl+C cancels)" — across en/zh/zh-TW, and the footer names
the composer's input mode, NORMAL in command mode and INSERT otherwise.
The cancelling line drops its "Ctrl+C exits" hint in vi mode, where Ctrl+C
re-cancels instead of exiting, and the approval, ask and plan card hints —
plus the plan choice list — drop their Esc mention there, where Esc is
ignored.
At the prompt, ^C with a non-empty prompt clears it and leaves shell mode,
in vi mode and out. In vi mode it first saves the trimmed draft to the
cmdline history when there is one, as send() does, so an accidental
interrupt never loses it (Up recalls it); a whitespace-only prompt is
cleared without being saved. In vi mode Esc no longer decides an approval
or ask card, where it is pressed out of habit; a row key or Ctrl+C decides
instead, and Ctrl+C still cancels a running turn even while an ask answer
is being typed.
[ui].commandmode is user-global only: a project reasonix.toml cannot set
it, so a cloned repo never rebinds the user's keys.
The fresh-start fallback was inferred from `--continue` being set, so a
bare `-r` picker run with `--continue` also skipped --copy even though the
picker resumes its choice in place and has nothing to duplicate. Carry an
explicit `startedFresh` out of session resolution instead, so only a
`--continue` that found no session skips --copy; the picker and `--copy`
alone stay usage errors.
`--continue --copy` with no saved session failed with "--copy requires
--resume or --continue", even though --continue had been passed. --continue
already falls back to a fresh session when it finds nothing, so there is
nothing to duplicate; --copy is now skipped and the run proceeds.
`--copy` with neither --resume nor --continue is still a usage error. Fixed on
both the run path (cli.go) and the TUI (tui_command.go).
The run path's copy resolution moves into applyRunCopy, which keeps runAgent
within its complexity ratchet.
OpenAI documents that GPT-6 Sol and Luna take function calls on Chat
Completions only at reasoning_effort=none, and that GPT-6 Astra's tool
calling requires Responses. The curated OpenAI preset nevertheless shipped
kind "openai" (Chat Completions) with default gpt-6-sol, effort medium and a
ladder without none, so every default turn was Chat Completions + tools +
effort medium: the one combination the model table (gpt6Capability,
Responses-only) already says cannot call tools.
- The preset's entry moves to kind "responses". Its effort list and default
are unchanged; both are inside every GPT-6 model's Responses ladder.
- shippedPresets gains the Chat Completions shape as a generation, with a
Kind field on shippedPreset: an entry still identified by preset_id
"openai", our base URL, the shipped catalog, window 1,050,000 and default
gpt-6-sol is moved onto the Responses wire. Anything its user changed in
that shape is left alone, and a second pass changes nothing.
- A consistency test walks every curated preset: a model whose table row
names its wires must be served on one of them, and declared efforts and
default must sit inside that model's ladder. The old preset fails it for
all three GPT-6 models.
- Effect tests through boot.Build hold a fresh install and an install of the
old shape to a Responses body carrying tools and reasoning.effort=medium.
Docs: https://developers.openai.com/api/docs/guides/latest-model,
https://developers.openai.com/api/docs/models/gpt-6-sol,
https://developers.openai.com/api/docs/models/gpt-6-luna,
https://developers.openai.com/api/docs/guides/reasoningCloses#11199
The agent's browser panel put its page away and showed "the page is hidden
while something covers it" when nothing a person could see was over it, and
sometimes stayed that way after the cover left.
Why it misfired:
- The check took the topmost element at nine points as the answer. A layer at
opacity 0, or a transparent full-window layer that only catches clicks, is
the topmost element and paints nothing, yet it hid the whole page.
- A cover that leaves by a CSS transition, an animation or a hover changes no
attribute the panel watched, so the page stayed away until something
unrelated moved. The observer also only listened to five attribute names.
- A popover reaching into one corner hid the whole page and replaced it with
a sentence.
What changes:
- ui/occlusion.ts decides from hit testing plus computed style and geometry:
an element counts only if it paints at the probe point (effective opacity,
visibility, background, backdrop filter, a border band, its own text rects,
or a replaced element). The slot's own content, its ancestors, and layers
that paint nothing are skipped. The rectangle sent to the shell is the part
of the slot the window shows, cut to clipping ancestors. A layer reaching no
more than 8px into the edge, like a column gutter's grip, does not count.
- Recovery no longer depends on a mutation: transition and animation
end events reschedule the check, all attributes are observed, and a 400ms poll
brings the live page back whatever moved.
- A covered page is frozen instead of hidden: the shell takes a picture of the
page on screen with capturePage before putting it away, and the panel draws
that picture in the slot until the cover leaves. The sentence is gone.
The picture channel (browser:freeze) takes only the page already shown in
this window, answers only this window's own page, and the picture is drawn in
an <img> and sent nowhere else. A freeze overtaken by a show does not put the
page away. An older shell without the verb falls back to hiding the page.
Tests: browser-panel.interaction.test.tsx covers invisible, transparent and
hidden layers, the slot's own content, recovery with no DOM change, stale
pictures and view transitions. test/freeze_live.js checks the channel in a
real Electron window.
A relay that serves several vendors' models under one address and key could
only declare one supported_efforts list for all of them. Each model can now
carry its own, stored where per-model settings already live
(model_overrides.<model>.supported_efforts / default_effort) and applied by
the same model resolution every request and every level picker reads, so a
level the picker offers is the level the request carries.
Precedence: the model's own list, then the connection's supported_efforts,
then the model table for the endpoint, then the protocol's generic ladder. A
model with no declaration inherits; installed configs need no migration.
- config: ModelEffortDeclaration / SetModelEffortDeclaration (keeps every
other override field, drops an emptied override, ErrModelDefaultEffortNotListed
for a default outside the levels) and InheritedEffortCapability. ResolveModel's
three branches share one forModel step.
- EffectiveEffort: the stored level is one per connection, so after a model
switch it can fall outside the next model's levels. That model now runs at
its own default, the level its picker's "auto" names, instead of omitting
the field while the picker says auto means something else.
- An override that only sets max_output_tokens is no longer dropped as empty
on load.
- serve: /providers lists each chat model's own, inherited and protocol
declarations; /providers/edit takes modelEfforts as the whole answer for the
listed models, refusing with provider.model_default_effort_not_listed and
provider.model_effort_unlisted.
- Studio: the "Thinking and reasoning levels" fold gains "Levels per model":
inherit (showing what that means) or pick the model's own levels and default.
The effect test builds a mixed relay through boot.Build and checks the
request body for every offered level of both models, and that a level a
model does not take never reaches the wire.
Closes#11192
The check that refuses a click as browser.covered answered a different
question from the one it was asked, in four ways:
- DOM.getNodeForLocation takes document coordinates, and it was given
viewport ones. On a page scrolled by less than the target's distance
from the top of the viewport it hit-tested a point that far above the
target and reported whatever sat there (a banner, a nav bar) as
covering it; scrolled further, the lookup found nothing and the check
passed, so a click under a fixed header or a veil went through to the
wrong element.
- ignorePointerEventsNone: true makes the hit test count
pointer-events:none elements as hits, the opposite of what a pointer
does, so decorative overlays were reported as covering.
- Newer Chromium (Edge 154) resolves a hit on a ::before/::after to a
CSSPseudoElement, which has no parentNode, so a button with a ripple
layer was refused as covered by its own pseudo-element.
- A form control drawn by its label (a zero-size or visually hidden
checkbox, a radio painted over by a span in its label) was refused as
not_visible or covered, although a person clicks the label.
The hit point is now converted to document coordinates, pointer-events:none
is left to the browser, a pseudo-element hit counts as its element, a hit
inside one of the target's labels counts unless it passes through other
interactive content (a link in the label follows the link), and before
refusing, the element is scrolled to the middle of the viewport, which
clears fixed headers and footers. A real cover is still refused, now
naming what is on top by ref. The label list comes from the page, so it is
capped and checked to be labels on the host side.
Measured with a table of pages at 1x and 1.5x device scale on Chromium 151
and Edge 154: before, 6 false refusals (7 on Edge), one click silently
landing on a fixed header and one passing through a veil; after, none,
and the true covers are still refused.
Every curated preset so far points at a Chinese coding plan or a
multi-model relay, so a user arriving with an OpenRouter, OpenAI or
Gemini key has to write the provider by hand.
- openrouter: DeepSeek, GPT-6, Gemini, Kimi, GLM, Qwen and MiniMax ids
checked against /api/v1/models; credits balance via the existing
OpenRouter decoder; HTTP-Referer / X-OpenRouter-Title /
X-OpenRouter-Categories so the traffic is attributed to Reasonix in
OpenRouter's app rankings. Claude ids are left out: on the openai
dialect no cache_control is sent, so every prompt token would bill at
the full input rate (claude-dialect).
- openai: GPT-6 Sol/Astra/Luna on Chat Completions, which already sends
max_completion_tokens to api.openai.com; declares the documented
low..max effort ladder.
- gemini: the Developer API's OpenAI-compatible endpoint.
- volcengine-coding-plan: the Coding Plan endpoint with ark-code-latest,
which follows the model chosen in the Ark console.
The presets live in their own file; provider_presets.go only appends
them.
A model can now declare optional run modes beside its effort ladder, in the
same model capability table. Each declaration names its id, the frontend's
label/hint keys, how it maps onto the request, the wires (kinds) and the
vendor endpoint it was established on, and whether it costs noticeably more.
GPT-5.6 (gpt-5.6 alias, sol/terra/luna) and GPT-6 (sol/luna/astra) declare
"pro", sent as Responses reasoning.mode "pro" beside reasoning.effort, on the
Responses API at api.openai.com only (developers.openai.com/api/docs/guides/
reasoning: mode and effort are independent; pro does more model work and is
billed at the model's rates for more tokens).
The mode is session state, not config and not prompt text:
- control.Controller lists the modes the session's model declares and sets
one; an undeclared mode is refused with ErrModelModeUnsupported.
- The agent carries the chosen id on every main-loop provider.Request; a new
conversation starts with it off. Subagents, compaction and side models
never carry it.
- The Responses adapter sends reasoning.mode only for a mode its endpoint
declared (projected by boot into the provider config), so a relay, Chat
Completions or a model without the declaration never sees the field.
- A rebuild carries the mode onto a model that declares it and drops it
otherwise, so a model switch turns Pro off.
- serve: POST /model-mode, refused as model_mode.unsupported; /status lists
modes only for a model that declares any.
Studio shows a switch at the bottom of the effort menu only when the model
declares a mode, marked as costing more; the trigger reads "... · Pro mode"
while it is on. The row stays pinned to the bottom of a scrolling menu.
Documents how Reasonix discovers and runs Agent Skills: the project,
custom, global and built-in roots, run_skill and slash invocation,
references/ and scripts/ handling, the extra frontmatter keys, and the
/skills and [skills] controls. agentskills.io asks listed clients to link
a page describing their implementation; this is that page.
Persist always-allow rules in private per-workspace user state and load them into the effective policy. Default skill installs to user scope unless project scope is requested, so neither action creates reasonix.toml in the workspace.
The model table named GPT-5.6's none..max ladder, but the provider was only
ever told a user-declared supported_efforts list, so on api.openai.com the
request builder refused "none" and "xhigh" at build time and clamped "max" to
"high". RequestEffortLevels now projects the resolved ladder (declared list
first, then the model table for the endpoint in hand) into the provider
config, so every rung the composer offers reaches the wire as itself.
Ladders, each from the vendor's own documentation:
- gpt-5.6 (alias of gpt-5.6-sol), sol/terra/luna: none|low|medium|high|xhigh|max,
default medium, on Chat Completions and Responses.
- gpt-6-sol / gpt-6-luna: none..max, default medium; gpt-6-astra: low..max
(none answers 400). Responses only: on Chat Completions GPT-6 takes
function calls only at reasoning_effort=none, and every turn carries tools.
- Model Studio Qwen3.8 (qwen3.8-max, -max-0902, -flash, -2.4t-a95b, -27b):
low|medium|xhigh plus none (enable_thinking=false), default xhigh; high and
max fold to xhigh, minimal to low. Vendor hosts: dashscope(-intl).aliyuncs.com
and <workspace>.<region>.maas.aliyuncs.com.
A ladder's vendor is now a host set with suffixes (vendorEndpoint), and a
ladder names the provider kinds it was declared for. The model table moves to
effort_models.go. The composer names none ("No reasoning") and tells xhigh
("Extra deep") apart from max, which shared one label.
reasoning.mode "pro" is not added here: it is a Responses-only dimension
independent of effort and needs its own control.
approval.go went 10 lines over repolint's 800-line ceiling with the MCP
origin handling. The ask-notification text builder now sits beside its
existing test file; no behaviour change.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The rewind menu's three scope labels and the bash coverage-gap reason are
Chinese literals held in a table and rendered as written:
<span className="lb">{s.label}</span>
Neither static guard could see them. The catalogue test only collects
Chinese inside a t() call, and the JSX test only reads text between tags or
in a rendered attribute. A table holding Chinese is correct exactly when its
render site calls t(), so the render site was the thing that was wrong. The
browser sweep that is meant to be the judge never opens this menu either:
tools/steps.mjs walks no rewind step.
Route both render sites through t() and add the four keys to the English
catalogue. A Chinese window is unchanged - zh still falls back to the key.
Verified: pnpm typecheck; pnpm test (169 files, 1035 tests).
Extract ProjectConfigPath(root) as the single place the project config
path is computed, and route every caller through it: load, credentials,
edit, retired-key migration, MCP migration and boot. It still resolves to
<root>/reasonix.toml, so behaviour is unchanged.
MigrateV2 rewrites a fact and then the index. When the index is unreadable,
reindexIn fails after the fact has already been rewritten, and the next run
skips facts that carry an id, so the entry stays out of the index for good.
The migration now reads the index once per directory before touching any fact,
the same guard RestoreArchived uses, so a failure happens before the first
write and a retry still sees the fact as unmigrated.
Cache note: no prefix bytes move; this only decides whether a migration starts.
Add a user-global [tools.shell] env table so every command the bash tool
runs inherits extra KEY=value variables (for example
`BASH_ENV = "~/.reasonix/bashsandbox.rc"`). The value is applied over the
inherited environment but under the host-owned session/egress overrides, so
a preset value never displaces TMPDIR or an egress token.
The variables ride the sandbox spec (sandbox.Spec.ShellEnv), so foreground,
background, and host-terminal runs all inherit them alike. It is
user-global: a cloned repo's reasonix.toml cannot inject variables, because
the shell env is held across the project merge next to the sandbox grants.
Presetting a variable that names a host endpoint (SSH_AUTH_SOCK, DOCKER_HOST,
CONTAINER_HOST, XDG_RUNTIME_DIR) does not unmask it: the denied authority set
resolves from both the host environment and the effective one, so a preset
that points a client elsewhere still leaves the host's own socket masked. The
egress loopback proxy ports are unioned the same way, so a preset proxy is
closed and overriding a variable cannot reopen the host's.
The interpreter itself is still resolved by [tools.shell] prefer/path. A key
that cannot appear in an environment entry is dropped rather than written as
a malformed one, and keys are sorted so identical config yields identical
environment bytes.
The recall diagnostic reported one Omitted count for two different decisions:
the per-turn hit cap and the character budget. The note the model reads already
names both, so a reader could not tell which one dropped a fact.
RecallResult now reports OmittedByLimit and OmittedByBudget apart, and the
/memory recall diagnostic prints both. The trajectory audit keeps its combined
Omitted value, so the persisted shape is unchanged, and the model-visible note
still carries the sum of the two.
From #11146's review: "Splitting omitted into limit vs budget would be a nice
separate diagnostic."
Address the two non-blocking nits from #11148's review: state the constraint
the test pins instead of narrating the implementation, and assert the error is
the pinned-budget refusal rather than any error.
The assertion matches on identity rather than on the message:
validatePinnedBudget wraps ErrPinnedBudget, and both the recovery test and the
save path's own test assert with errors.Is. The sentinel's text is the setting
the message already names, so the user-visible message stays byte-identical.
Between roughly 470px and 720px of conversation column (a wide browser
pane is the usual way there) every toolbar control was flex: none, so the
model/effort group ran past the toolbar's box: the send or stop button sat
on the effort picker and the effort label painted outside the composer.
Below 470px the send cluster was taken out of flow with a fixed 38px
reserve, which the wider stop button does not fit in.
The row now lays out in flow at every width: the send cluster keeps its
own slot at the end of the line, the toolbar wraps inside what is left,
and the model group has a minimum basis, grows to fill the line and lets
the model name shrink to an ellipsis before it moves to a line of its
own. The spacer and the per-width offsets it needed are gone.
perf/composer.mjs sweeps 340-760px idle and running with a long model
name and fails on any control outside the composer, any two controls
overlapping, or a model name narrower than 40px; it fails on the old
stylesheet at 340px (running) and 480-720px (both states).
Fixes#11159
A run that keeps calling tools without changing anything the host can
observe (#11153: dozens of successful `echo` calls with varying arguments,
about 1.1M input tokens) had no signal short of reading the transcript.
This adds one, without bringing back any round-count guard: nothing is
sent to the model, no call is refused, and the cache-stable prefix and
the per-turn execution-policy block are untouched.
Judgement (kernel, structure only):
- evidence.EffectWatch scores each round's receipts. An effect is a
successful mutation or named write, a task step moving (StepProof,
TodoStep, or the host's todo progress counter), a verification whose
outcome changed since it last ran, a first read (a read-only tool
answering a request not seen before in this run, or a shell command
naming an existing path not named before, resolved by stat), or a
delegated result coming back. A command merely being new is not.
- agent.settleProgressRound runs at each tool-round boundary; the count
lives in turnRuntime, so a user message zeroes it. A second,
model-independent backstop compares the input tokens spent since the
last round with an effect against context window x N; an effect
restarts both counts, so a long run that keeps doing observable work
is never flagged however much it spends. Without a ledger the round
judgement stays silent rather than guess.
Feature 2 (default on): a content-free progress_watch event carries the
reading; Studio shows a one-line strip above the composer with Stop
(the ordinary cancel) and "don't show again (this session)". It clears
itself when an effect lands or the turn ends. The TUI says it once per
stall.
Feature 3 (default off): [progress_watch] pause = true ends the run at
the round boundary through a resumable pause (RunPauseInfo Kind
no_progress, TurnDone outcome no_progress). No finalization nudge is
injected. Continue sends an ordinary message, which starts a new run
with the count at zero. The section is user-level only: load restores
the user's value over a project reasonix.toml. Settings > Session edits
it through GET/POST /progress-watch, which persists to the user file
and hands the new watch to the running executor without a rebuild.
Out-of-range values refuse with progress_watch.out_of_range.
Refs #11153
The market entry gains up/down vote buttons that show the signed-in
account's vote; pressing the lit one withdraws it. Signed out, the entry
offers sign-in (a jump to the account section); a publisher, an
unverified account and a paired device each say why they cannot vote.
List rows show approval and install count, and the listing defaults to
the registry's new recommended order.
The kernel owns the calls: market.Client gains MyVote/Vote, which send
the account token only to the fixed registry host with redirects
refused, and ReportInstall. Registry refusals keep their identity
(signed_out, email_unverified, own_package, rate_limited; an identity
outage stays market.unreachable) through market.* refusal codes. Vote routes sit
behind accountAuthAllowed like every other account route.
A successful market install reports once, carrying only the slug and a
random id kept only for the market (market/install-id under the Reasonix
home, unrelated to the usage-statistics id), and only when the host allows
it (release builds), both desktop statistics switches are on, and
neither DO_NOT_TRACK, REASONIX_TELEMETRY=off nor CI opts out.
The [desktop] telemetry and metrics switch descriptions now say that
per-package market install counts are reported (anonymous, no content).
A publisher's own unreviewed install is never reported.
The usage panel summed every turn in the currency its price table was
kept in, and ignored the display currency the rest of Studio reads
costs in. A DeepSeek entry holds the vendor's USD table, so a user who
pinned CNY, or whose CNY wallet selected it, saw "US$" on the panel
while the pane's cost meter said "¥".
Every quote already carries the vendor's own table in both currencies,
and the stats line records both valuations; nothing had to be converted,
only read. The usage catalog now keeps each row's valuations and tallies
one partition per view: a row the vendor priced in the display currency
moves into it, and any other row keeps its billed currency, so a relay's
USD price is still shown as USD rather than converted at an invented
rate. The schema version moves to 5, which rebuilds the projection from
the JSONL it is derived from.
The live projection path also dropped the cost of every row a running
process appended, marking the file indexed all the same, so today's
spend vanished from the panel until the next restart. It now carries
the same cost a reindex would.
A skill in the Studio composer was only the text "/name", and only a
leading one ran: the kernel reads a slash at the start of a line and
nowhere else, so "/review" typed mid-sentence reached the model as prose.
Picking a skill or subagent from the slash menu now places a chip, at the
start of the line or after any "/" that opens a word. A "/" past the start
asks the kernel's completion as if it opened the line and keeps only
skills and subagents, since a command is not something a sentence can
hold. The chip is the record of the pick; the same letters typed by hand
stay prose, so nothing reads the submitted wording to find a skill.
The chip is drawn by a mirror laid over the textarea at the textarea's own
computed metrics, so caret, selection, IME and undo stay native. Edits are
reconciled from the text the chips were last placed in, so every way into
the box (typing, drops, prompt refine, a refused send) carries them, and an
edit that reaches into a chip removes it whole. Skills take --net and
subagents --deleg.
On send the line goes out as typed (input), the text with the chips taken
out (submit), and the chips (invocations). /submit and POST /inbox/items
route a chip line through the inbox envelope, which already carries
invocations to prepareInvocationTurn, so a mid-line chip reaches the model
byte-identical to a leading "/name task". A kernel that predates this reads
only input and still runs a leading chip. TryEnqueueAndSteer queues a line
holding an invocation as a follow-up: a steer is read as guidance and would
drop the skill.
Closes#11145
RestoreArchived checked id and name collisions but skipped the rest of what a
save validates: a recovered fact could take a subject_key another active fact
now holds, or push pinned guidance over the configured budget. The fact model
states the subject rule unconditionally, and the save path refuses both.
Validate the restored fact with the same validator SaveWithOptions uses, before
any part of the restore is written.
studio still drops the read error when it rewrites the MEMORY.md index:
store.go reads the existing index for indexLinesExceptIn and flushIndexIn
with `existing, _ :=`, so an unreadable index reads as empty and the next
save rebuilds it from the managed lines alone, silently discarding
hand-written content while reporting success.
main-v2 already fixed this in 2eb878989 ("preserve index when reads fail"),
with 79acaf6e9 moving the readers into store_index.go to stay inside the
repolint budgets. This ports the same shape to studio so both lines behave
alike:
- readIndexIn keeps treating a missing index as empty but returns any other
read error; indexLinesExceptIn reports that error and whether the name was
present, which lets indexContainsIn go away.
- flushIndexIn and reindexIn propagate the read error instead of rebuilding
the index from an empty read.
- archiveLocked and archiveMemoryInDir read the index before archiving, so a
failed read happens before any side effect.
- SaveWithOptions reuses removeMemoryFromDir and archiveMemoryInDir, dropping
the duplicated remove/flush and archive/flush pairs and keeping the
function inside its size and complexity budgets.
AppendDoc discarded the read error, so an existing but unreadable doc-memory
file (REASONIX.md/AGENTS.md) was treated as empty and rebuilt from scratch,
silently destroying the user's standing instructions. Surface the read error
unless the file is simply missing.
Ported from 1.x (#8319), which fixed the same defect on main-v2; the 2.x
writer still had the original discard.
Entries may carry a discussion number instead of an issue number. Discussions
have no REST comment API, so the reply goes through addDiscussionComment and
skips a body the bot already posted there.
A publisher can now keep a package to themselves and install their own
packages before, or without, review.
- Publishing takes a visibility. "Only me" saves the package as private:
it skips review and never appears in the market. My packages marks it
private and offers "Submit for review", which moves it to pending.
- Every row in My packages can be installed. The kernel reads the package
from the registry's /v1/me/packages/:handle/:name as the signed-in
account, never from what the window sends, so another account's package
is refused (market.not_yours).
- A live version keeps its reviewer's pin. Any other version has none, so
the preview's own contentDigest becomes the pin: apply must echo it
(market.unpreviewed otherwise) and install_source refuses material that
changed since the preview. The confirmation says "Unreviewed, visible
only to you", and the ledger records the install as unreviewed.
- The plan, the risk grades, the multi-skill confirmation and the
themes-only check are the same ones every market install passes; the
confirmation view is shared with the market entry page.
Each market row is now two lines: name, kind, verified/installed badges and
the install count on the first; a one-line truncated summary with the
publisher and version, de-emphasised, on the second. Rows stay buttons, so
they open with the keyboard, and truncate rather than overflow at phone width.
The listing now says which packages can be installed from the market. The
kernel reads the registry's `pinned` per listed package (nil when an older
registry does not say, so nothing is claimed) and forwards `pinned=1` for the
new "Installable only" toggle. The registry applies that filter, so pages stay
whole; a registry that answers a filtered ask with an unfiltered page is
refused as market.filter_unsupported rather than trimmed client-side. The
install path is unchanged and still checks the approved version's digest.
The chunked update stages the new tree in the update cache and a helper
swaps it in after exit, renaming each installed file into a backup
directory beside the staged tree. Two installs can never complete that
first rename: an all-users install under Program Files, which grants its
users read and execute only, and an install on another volume than the
user's cache, where MoveFileEx without MOVEFILE_COPY_ALLOWED refuses.
For both, the helper rolled every file back and relaunched the old build
with nothing shown, so the update downloaded, "installed", and the
application came back on the version it started from. Before chunked
updates these installs took the full installer, which raises its own
consent and writes wherever it installed.
tryDelta now proves that move before fetching anything: it creates an
entry in the install root and renames it into the cache's delta
directory. If either step fails the full package is downloaded instead,
as for any other delta that cannot be used.
Closes#11115
In a conversation column of 460px or less, app.css stretches the send
cluster to width: 100%, and at 470px or less studio.css takes the same
element out of flow (position: absolute; right: 0; bottom: 0). The two
rules combined into an invisible full-width strip over the composer's
last line, so every tap on the model picker or the effort picker landed
on the send cluster. The pickers rendered but never opened on a phone,
over LAN and over the relay alike.
The absolute rule now declares width: auto, so the cluster is only as
wide as the send button. A stylesheet test resolves the cascade for
the cluster at phone widths and fails if it is out of flow and a full
row wide.
Closes#11116
An over-budget worktrees/ or modules/ group was bound read-only as a whole,
but the exact mounts for the gitdirs inside it were still emitted first, so
planting submodule gitdirs kept growing the bubblewrap argument list. Paths
beneath a read-only tree are now left to that tree.
The serve launch-token sandbox test placed its files under /tmp outside the
write root, which bubblewrap masks with a tmpfs: the control read failed and
the denied reads passed without exercising the forbid-read mounts. The files
now live inside the write root on every platform.
Add English wording for the serve.host_rejected and
auth.launch_token_required refusals.
Signed-in users can now submit skills, plugins, MCP servers and themes to
the community registry from Settings > Extensions > Discover, and follow
each submission's review state.
- internal/ext/market: Client.Publish (POST /v1/packages) and Client.Mine
(GET /v1/me/packages) send the account session as a Bearer token to the
same fixed https host the listing reads, with redirects refused, so the
token never reaches another host. The registry's refusal codes are
projected onto sentinels (ErrSignedOut, ErrEmailUnverified, ErrNotOwner,
ErrVersionExists, ErrRateLimited, ErrRejected).
- A submission is checked against the same source rule an approved
version must pass to install (sourceInstallable, shared with
installable): plugins and themes must name a GitHub tree pinned to a
full commit. Nothing is sent for review that could never be installed.
- theme is its own listing kind that installs through the plugin
installer (Installer("theme") == "plugin") under the same commit pin and
digest check. The plan must consist only of plugin packages that carry
themes and nothing else (no skills, commands, agents, hooks, MCP servers,
prompts or runtime); otherwise market.not_theme.
- serve: POST /market/publish and GET /market/mine, gated like every
account route (host grant plus a stored session), each failure a dotted
code (market.signed_out, market.unpublishable, market.rejected, ...).
- Frontend: Browse / My packages / Publish views inside the market group
while signed in, a sign-in prompt otherwise; a Theme category filter.
- A theme's apply re-plans and checks themes-only before writing, because
a planId binds the source and actions but not the listing kind. A
registry accounts_unavailable answer reads as market.unreachable, not
as signed out.
The host's own git reads the workspace repository's configuration and hooks,
so a confined command that can rewrite them can make the host run a program
of its choosing. The OS sandbox now denies writes to them while leaving the
rest of .git writable.
For the repository git itself discovers from each writable root, protected
paths are: config, config.worktree, commondir and hooks/ of the gitdir and
of the common dir; config, config.worktree and commondir of every existing
linked worktree entry (new entries stay creatable); config, config.worktree,
commondir and hooks/ of every submodule gitdir under modules/. A .git file is followed to the gitdir it names
exactly as git resolves it, and the pointer file and commondir are
protected, so a repository cannot aim the protection at a decoy while git
reads something else. The .git entry, the gitdir, the common dir, every
symlink on the way to them and each writable ancestor are pinned: they
cannot be removed, renamed away or replaced by a symlink, while their own
mode and times stay writable (touch and chmod on the workspace root work).
- macOS: deny rules after every write allowance. Existing worktree and
submodule gitdirs (up to 128 each, found without following symlinks) get
exact rules and pins; patterns anchored at the common dir cover the rest
and gitdirs created later, skipping refs/ and logs/ segments so a branch
or tag named config or hooks stays writable. The profile stops growing
past the limit.
- Linux: pinned directories become their own bind mounts and protected files
and hook trees are bound read-only; only real directories and regular
files are mount targets. Up to 512 gitdirs per group get exact mounts;
past that the whole group directory is bound read-only, so planted HEAD
files cannot overflow the argument list. Only existing paths can be
mounted and symlinks cannot be pinned.
- Windows has no Bash sandbox; the guide says so.
A protected file with another hard link refuses the confined launch with
sandbox.git_metadata_linked (ErrGitMetadataLinked), since no path rule sees
a write through the other name. Whenever a confined command's output names
a protected path, also on exit 0 (git reports some refused config writes
and exits 0), the bash result carries sandbox.git_metadata_protected.
Real-sandbox tests cover each protected path, pointer and symlink
retargeting, patterns for gitdirs created later, bounded rule size with
2000 submodule gitdirs, ordinary git work, and the git operations that now
fail or silently skip their config write.
Every git invocation the host makes goes through gitcmd, and each one must
leave the programs a repository's own configuration names unexecuted:
filter and merge drivers, textconv and external diff, fsmonitor, hooks,
signature verifiers, submodule-boundary programs and lazy-fetch transport.
- Fixed keys are overridden with -c: core.fsmonitor (empty, not "false",
which git before 2.35.2 reads as a hook name), core.hooksPath,
log.showSignature, merge.verifySignatures, diff.submodule, gc.auto and
maintenance.auto and safe.bareRepository=explicit; submodule.recurse=false
for checkout, switch, restore, reset, merge and read-tree (host tree
updates never update submodules), and elsewhere only where the
repository's own config sets it.
GIT_NO_LAZY_FETCH=1 stops a missing object becoming a fetch through the
repository's remote (git 2.44+); a historical patch that fails for that
reason returns gitcmd.ErrObjectNotLocal, decided by git's own
--missing=print listing.
- diff, log and show carry --no-ext-diff and --no-textconv; status and diff
carry --ignore-submodules=dirty. The subcommand is found past global
options, so a call passing -C inside args is hardened too.
- Filter and merge drivers defined at local or worktree scope are listed by
git itself before each invocation (includes followed) and overridden empty;
user and system scope stay live; a name -c cannot address fails closed
(ErrRepositoryDrivers).
- worktree add uses --no-checkout and populates through gitcmd inside the new
worktree, so a conditional include keyed on the new gitdir or branch is
seen by the driver listing there rather than run by the add's checkout.
The reset takes the worktree-mutation timeout, and a failed reset removes
both the worktree and the branch the add created.
- Launcher ls-remote runs Detached with gitcmd's environment (only PATH is
taken from the launcher's declared env), requires an https://, ssh:// or
git:// remote (case-insensitive; http:// and local paths are refused), and
passes the remote and ref past -- so a leading-dash value is an operand.
- frontend/cli runGit passes its directory through gitcmd so enumeration runs
in the target repository, not the process cwd.
- Repository identity is resolved once, when a session opens: boot runs
gitcmd.Open (rev-parse --absolute-git-dir --git-common-dir --show-toplevel)
before any command, the controller carries the resulting gitcmd.Repo, and a
rebuild of the same workspace passes it on. Every later host call runs
through the Repo with GIT_DIR, GIT_COMMON_DIR and GIT_WORK_TREE set, so a
nested .git, a commondir file or a damaged HEAD written afterwards cannot
substitute another repository or its config; a damaged HEAD fails closed.
Threaded through gitstatus Status/Diff (serve /changes and diff),
worktree Inspect/Create/TakeSnapshot and the candidate/apply/remove paths
(Snapshot and Candidate carry their Repo; host-created worktrees are opened
right after worktree add), the isolation store, best_of_n, and cli review
(resolved once before the review agent runs; ObjectsMissing is a Repo
method).
- Open accepts only a git dir belonging to the work tree that holds the
directory: the work tree's own .git (sharing no other common dir) or a
linked worktree's admin dir under <common>/worktrees whose gitdir file
names that work tree's .git. A .git file or core.worktree pointing at
another checkout is ErrNotRepository. Submodule checkouts and
--separate-git-dir repositories opened directly are therefore treated as
not version-controlled.
- Repo.Verify applies Open's test to a stored identity; Snapshot and
Candidate identities are verified when used again.
- The snapshot's private index copy keeps the real index's mtime: with a
fresh mtime git trusted stat for entries written in the index's own
second, so a same-size edit made in that second was left out.
- Snapshot staging uses gitcmd Repo.StageAll: gitlinks present as directories
and untracked nested repositories are excluded with :(exclude,literal)
pathspecs, because add -A checks a submodule by running git inside it under
its own config. A gitlink keeps the commit the index records; an untracked
nested repository is left out (it previously failed the snapshot when it had
no commit).
Regression tests use real git with payload scripts: the gitcmd config-key
matrix (signatures, lazy fetch, submodule boundary), the worktree call sites
(conditional includes on Create and CreateCandidate), the launcher's non-URL
remote rejection, identity pinning after a nested .git, a commondir file and
a damaged HEAD (config origin and filter marker), Open refusing another
checkout's git dir (.git file, core.worktree, foreign admin dir), StageAll
with nested repositories named like pathspec magic or globs, StageAll and TakeSnapshot
past a gitlink with its own filter, the listing and snapshot of a workspace
that later gains its own .git, and a global submodule.recurse=true against
host checkout and reset. The list-then-run race is left to a follow-up that
denies sandboxed writes to .git config and hooks.
`reasonix serve` now generates a per-launch token unless `--auth` or
`[serve].auth_mode` chooses otherwise. `none` remains an explicit opt-out for
deployments fronted by their own authentication: reads stay open to loopback
and bound names only (421 `serve.host_rejected` otherwise), and every
state-changing request, approvals included, needs the launch token (bearer
header or the /auth/token cookie) and otherwise refuses with
`auth.launch_token_required`. A request the loopback gate, a paired-device
gate or the in-process transport already admitted counts as holding it, so
Studio and the TUI keep working unchanged. Token mode also accepts the bearer
header.
The token never reaches the terminal: serve writes it to a 0600 file under the
remote state directory and prints the path, or names the --token-file of a
managed launch. Runtime sandboxes and read tools deny that directory and every
`--token-file`, `--provider-broker-file` and `--provider-broker-token-file`.
[serve] is read from the user config only, so a project reasonix.toml cannot
choose the token, the auth mode, or trust in forwarded headers.
Studio can now browse, search and install from the community registry
(Settings > Extensions > Discover). The market is a source, not a trust root:
- internal/ext/market reads the registry read-only: https to the fixed
crash.reasonix.io host, 15s timeout, capped bodies, redirects refused,
non-active rows dropped. Registry text is untrusted data.
- Only the approved version is installable, and only when its row carries a
content digest. The digest travels to install_source as expectDigest.
- install_source gains a contentDigest on every plan (machine-independent:
SKILL.md bytes, plugin commit, MCP entry) and refuses a plan whose material
differs (ErrDigestMismatch) or cannot be pinned (ErrNotPinnable). Preview
and apply both pass the check, and apply writes what its own plan read.
- GitHub plugin sources pinned to a full commit (tree/<sha>) now clone and
check out that commit instead of failing on --branch <sha>.
- A remote source that expands into more than one skill is graded high, so
the permission layer asks a person even under a blanket allow, and the
market confirmation lists every skill and holds the install button until
the person ticks that they read them.
- Market installs require the planId of the plan that was shown, go to the
global scope, and are recorded in <home>/market/installed.json; "installed"
is read back only while the recorded targets still exist.
- serve: GET /market/packages, GET /market/packages/{handle}/{name},
POST /market/plan, POST /market/install, each failure a dotted code
(market.unpinned, market.content_changed, market.unreachable, ...).
Adds an opt-in cloud backup under Settings > Account, shown only while
signed in. The kernel collects the user-level setup by category, seals
it locally (gzip, then XChaCha20-Poly1305 under an Argon2id key derived
from a passphrase the user sets), and uploads only the envelope to the
accounts service; the service never sees plaintext.
Categories: settings (providers, default model, interface look),
extensions (user skills, plugin references by source and version, MCP
servers), memory (user-level instruction files and global facts),
automation (global hooks, status line) and secrets (stored provider
keys, off by default). Without secrets, literal env/header values are
stripped with the plugin exporter's StripCredentials and URL userinfo
is dropped; the account session token never travels. Approval mode,
permissions and sandbox are deliberately not restorable.
Restore is preview-then-apply. The preview decrypts, diffs each item
against this machine, shows everything an item would write (endpoints,
env, headers, cwd, instruction file bodies) and holds the snapshot under
a single-use plan id. Apply recomputes consent against the machine at
apply time and refuses, writing nothing, unless each of these is named
in `consented`: hooks, the status line and MCP servers (code that runs
here); a provider that is new or changes endpoint, and a default model
pointing at one (conversations and keys go there); a key that replaces
one already stored; an instruction file that imports other files.
A snapshot is rejected outright when a key is not read by a provider it
carries, when a provider reads REASONIX_* or a host-steering variable
(proxy, TLS, loader, PATH), or when a unit's payload names another item.
Plugins are never installed by restore; they come back as sources for
the ordinary plugin install flow and its plan approval. File items are
written only under their named root, through validated relative paths
that may not cross a symlink; hidden files are not exported. Config
writes go through config.EditConfigFile, hooks through hook.Save, keys
through config.SetCredential.
Serve: GET/POST /backups, DELETE /backups/{id}, POST /backups/{id}/preview,
POST /backups/apply, gated by the account grant and a signed-in token,
with dotted refusal codes the frontend words in both languages.
The hosted build has no kernel behind its origin, so opening
studio.reasonix.io directly probed /runtimes, got the SPA shell and
reported that the kernel did not answer. The deploy now declares where a
visitor picks a computer, and a visit without ?device goes there.
A plan whose steps name candidate files owes a change the host proved.
The shadow verdict and the receipt already said so; the gate now holds
an answer that changed nothing, and releases a turn that concluded
blocked, is waiting on the user, or carries a change a restart brought
in. The hold names all three ways out.
The executor-handoff nudge now runs before the gate ends the run: an
executor that has not acted yet gets its one in-turn nudge first, and
the gate holds only what the nudge could not recover.
publish downloads studio-windows-signed by name, but any job in the run can
upload an artifact. windows-sign-installer now outputs the SHA-256 of the
installer it signed and verified, and publish requires the downloaded
artifact to hold exactly the installer and the portable zip, hashing to
that output and to windows-verify-package's archive hash. A mismatch, a
missing hash or an extra file fails with a studio-signing.* title before
anything is signed or published.
The Microsoft-signed d3dcompiler_47.dll and dxil.dll must now also chain to
Microsoft Root Certificate Authority 2010, pinned by thumbprint; subject and
issuer names alone can be issued under any trusted root.
The plan-criteria gate and the frozen contract verdict measured "passed
after the latest change" from LatestSuccessfulMutationIndex, which counts
an unclassified command as a change. A plan check such as `python -c ...`
is exactly such a command, so it became the latest change and owed itself
a later run that could never come: every plan whose checks were inline
scripts ended with its criteria outstanding, graded pass or not.
Both now read LatestProvenMutationIndex, the baseline project checks and
owed review already use: the latest proven write, or the latest unproven
one only when nothing was proven.
The Windows release signed only the three executables Studio builds. The
installed tree also ships Electron's ffmpeg.dll, dxcompiler.dll,
vk_swiftshader.dll and vulkan-1.dll unsigned, and Smart App Control / WDAC
can refuse to load an unsigned low-reputation DLL, so the app may not start.
Every PE file the Windows bundle may hold is now declared in
scripts/windows-signing-lib.ps1: seven the release signs, and
d3dcompiler_47.dll and dxil.dll, which Electron ships signed by Microsoft
and which are kept that way and verified as Microsoft-signed. A bundle with
an undeclared or missing PE file (.exe, .dll, .node, or any MZ header) fails
until a person updates the list.
windows-sign-payload signs the declared files, verifies all nine, and
records a SHA-256 digest of the whole signed tree as a job output. The
signed tree is the payload artifact; windows-package packages it as it
came. A new job, windows-verify-package, with no secrets and no
environment, checks the payload against that digest, unpacks the zip and
the installer's app-64.7z, requires both trees to equal the payload file
for file (plus resources/elevate.exe, pinned by SHA-256), and outputs the
SHA-256 of both packages. windows-sign-installer only requires the packages
to hash to those values before signing, and afterwards only runs SignTool
verification, so no archive is parsed in a job holding the Certum
credentials and Certum never signs an installer whose contents did not
match.
Signatures are read from the file itself through SignTool and the embedded
signer certificate. Get-AuthenticodeSignature answers from the OS catalog
for d3dcompiler_47.dll and reports a catalog signature whatever the file
carries.
The contract pins the new scripts, and validate now refuses a fingerprint
file that runs a script under scripts/ which is not pinned.
Root cause: some Anthropic-compatible gateways, DeepSeek's among them, open
the web_search_tool_result block with an empty content array and send the
results in a later web_search_tool_result_delta. The stream decoder declared
the delta's results field but never handled the event, so the search produced
no provider-run call: no card, and the only visible source list was whatever
the model wrote into its answer. 1.x fixed the same gap in #9197.
Fix: the server-run search state (serverBlock) now remembers result blocks
that opened empty and completes them from the delta, once per block. The
block-start and argument handling moved onto serverBlock with it, which keeps
readStream within its size and complexity budget.
Test: TestStreamSurfacesWebSearchResultDelta fails on origin/studio
("want 1 provider-run call from the result delta, got 0") and passes after
the fix; go test ./internal/model/anthropic/ passes on macOS and Windows.
Fixes#8900
Settings > Appearance > open theme folder failed on Windows with
`exec: "explorer.exe": executable file not found in %PATH%`.
Root cause: revealFolder started a bare "explorer.exe", which exec
resolves through PATH. explorer.exe lives only in %SystemRoot%, and the
environment Studio is launched with does not always carry that directory
on PATH, so the lookup failed before anything started.
Fix: resolve explorer.exe from %SystemRoot% (falling back to %windir%),
the same way the 1.x desktop's revealPath already does, and keep the
bare name only when neither variable is set.
Test: TestFolderCommandFindsExplorerWithoutPath sets PATH to an empty
directory and asserts the command resolves to %SystemRoot%\explorer.exe.
On Windows it failed on the base with the reporter's exact error and
passes after the fix; the serve Folder/Theme tests pass.
Fixes#11079
ACP clients render their token/cost meter from the stable usage_update
session notification (Session Context Size and Cost). The ACP server
never sent it, so clients such as Paseo showed no usage at all even
though the controller already measures the context gauge and the
session status telemetry already totals cost.
After every prompt turn, and after session/load and session/resume,
the server now sends usage_update with used/size from the controller's
ContextSnapshot (the same measurement the compaction trigger uses) and,
when every priced event produced a quote, the cumulative session cost
with its ISO currency. A session with no context window yet sends
nothing, since used and size are required by the protocol.
Test: TestPromptTurnReportsUsageUpdate drives session/prompt through the
real controller and agent with a scripted provider; it fails on the base
("usage_update notifications = 0, want 1 per turn") and passes with the
change. TestUsageUpdateForCarriesCostOnlyWhenPriced pins the wire shape.
Root cause: the live transcript draws a question from ask_request and
hides the ask tool call, since both describe the same prompt. A reload
has no event stream, so fromHistory rebuilt only the tool call - which
the transcript then hid. An answered question vanished entirely after a
reload, session switch or restart, even though its arguments (questions,
options, descriptions) and its result (the recorded answer) are both in
the history.
Fix: when an ask call has a successful result, fromHistory rebuilds it
as a sealed ask card from the call's own arguments, shaped the way
agent.AskTool shapes them for the wire, and carries the result verbatim
as the recorded answer. The card shows every question and option and,
in place of the per-question picks it cannot know, what the kernel
recorded the user said. A call with no result or a failed result stays
a plain tool call. The recorded answer is also searchable.
Test: src/state/ask_history.test.ts and
src/ui/cards/AskCard.recorded.test.tsx fail on the base (no ask item
after fromHistory; the footer reads "Library：未答") and pass with the
fix. Full frontend-next vitest 988/988, tsc clean, make check clean.
TestStreamSurfacesEarlyConnResetAsInterrupt and
TestStreamDoesNotReplayAfterOutput forced a real TCP RST right after
writing the 200 head. Whether net/http observes that RST before or after
the head is up to the OS scheduler. When it lands first, the failure is a
header-phase connection error, which SendWithRetry retries by design, so
the server counts several requests and the reqs == 1 assertion fails
(#8327). A local probe that resets before the head shows the path: the
server saw 11 requests and the call spent 92s in backoff.
The fixture is now a scripted RoundTripper: a 200 SSE response whose body
yields the prelude and then fails with a net.OpError(ECONNRESET). The
phase is fixed, the request counter is atomic, and no socket is involved.
Header-phase retry keeps its own coverage in provider's retry_test.go
(TestTransientErr, TestSendWithRetryRecoversAndNotifies).
Test: both tests pass 500/500 on Windows and on macOS. Mutating ReadCut
and streamFailure so a reset is no longer a StreamInterruptedError makes
both fail, so the fixture still guards the contract.
A Windows launch from a shortcut has no console, so every way Studio
could fail to start -- boot() throwing, the host exiting before its
handshake, the host never starting -- ended in app.quit() with nothing
on screen and nothing on disk.
The shell now writes <userData>/logs/shell.log (start, lock refusal,
host start/exit/spawn failure, boot failure with stack) and tees the
host's stderr into host.log, each rotated at 1 MB x 3. The launch
credential and any token-, secret- or password-named value are
redacted before reaching either file. A launch that fails shows an
error box (zh/en by the system locale) with the reason, the host's
last stderr lines when it never handshook, and the log folder, then
quits. A second launch refused the instance lock stays silent and only
leaves a log line.
Models often wrap an address in backticks ("open `http://127.0.0.1:8787`").
Markdown links and bare URLs were clickable in the transcript, but a code
span had no link handling, so the address could only be copied out.
Root cause: GFM autolinking never looks inside inline code, and nothing
else turned a code span into a link.
Fix: a remark plugin rewrites an mdast inlineCode node into a link when its
whole value parses as one http(s) URL with a host. mdast keeps inline code
apart from fenced code, so listings are untouched; a span already inside a
link is left alone. The existing click routing (closest a[href]) opens it
like any other link, and the code inherits the link colour.
Test: Markdown.plugins.test.tsx gains two cases. "links a code span that
holds exactly one web address" fails on origin/studio (expected null to be
truthy) and passes after; the negative case (command with a URL, file://,
bare scheme, fenced block) stays unlinked. tsc and the full frontend-next
vitest suite (155 files, 985 tests) pass; make check is clean. Real-browser
screenshots of the perf build, before and after, light and dark, show the
span underlined in the link colour and the fenced block unchanged.
Problem: files the agent creates or deletes appear in the Studio explorer
only once the whole turn ends (or the window regains focus), so a long
turn shows a stale tree for its full duration.
Root cause: the pane re-read the working-tree change list only on turn
boundaries, and the explorer reloaded its listing only when that list,
the running flag, the query or the focus glance changed. Nothing between
turn start and turn end asked the disk again.
Fix: count the finished tool calls that are not read-only (the kernel's
own readOnly declaration on each call) and use that count as a trigger:
the pane re-reads /changes and the explorer re-lists its tree and the
open file whenever it moves. Read-only calls cost nothing extra.
Test: pane-tree-refresh.interaction.test.tsx asserts /changes is read
again after a write_file result and not after a read_file result;
WorkbenchPanel.test.tsx asserts the tree picks up a new file when the
count moves mid-turn. Both fail on origin/studio ("expected changes to be
called 1 times, but got 0 times"; draft.txt never listed) and pass with
the fix. tsc --noEmit and repolint -changed pass; the full frontend-next
vitest run passes 982 of 983, the one failure being a 5s timeout in
session-new.interaction.test.tsx under full-suite load that passes when
run alone, with and without this change.
Fixes#7338
An edit whose old_string drops (or adds) a blank line the file has between
two otherwise identical lines never matches: the fuzzy matcher frees only a
blank-line run's length, never its presence. The failure then read as
"not found (nearest line N)" plus "re-read the file, check for stale
context", so the model re-read the file and reproduced the same span the
same way, over and over (#6337).
The host can tell this case apart structurally: with blank lines dropped on
both sides, old_string names exactly one span, and that span's blank-line
layout differs from old_string's. oldStringNotFoundError now reports that
span and its cause. Matching is unchanged; nothing new is accepted.
Test: TestEditFileNotFoundNamesBlankLinePresenceDrift (edit_file and
multi_edit, CRLF file) fails on studio with the generic nearest-line
message and passes after; two negative cases hold the claim to one span
with a genuinely different layout.
Fixes#6337
desktop/ replaces reasonix with the parent module, so the root bumps raise
its graph too, and the Studio desktop and package jobs failed with
"updates to go.mod needed". go mod tidy in desktop/ moves only the
versions the root already requires. This also covers #11033.
Model requests on the OpenAI Chat Completions, Anthropic Messages and
Responses wires, and the /models listing, carried no User-Agent, so Go's
default Go-http-client/1.1 went out. Gateways and relays in front of an
endpoint could not attribute the traffic, and some risk controls aimed at
generic Go clients blocked it.
provider.ApplyClientIdentity now sets User-Agent: Reasonix/<version> on
those requests unless one is already present, so a configured custom
header and the OpenCode Go identity still win. Both hosts (cmd/reasonix
and cmd/reasonix-studio-host) record their ldflags version once at start
through provider.SetClientVersion; an unversioned build reports "dev".
Test: TestRequestsIdentifyReasonix (chat + model listing),
TestMessagesIdentifyReasonix and TestResponsesIdentifyReasonix record the
User-Agent an httptest endpoint receives. Before the wiring all three saw
"Go-http-client/1.1" and failed; after it they see "Reasonix/9.8.7".
TestConfiguredUserAgentWins keeps a custom User-Agent header; the existing
OpenCode Go tests still require exactly "Reasonix" there.
Root cause: on DeepSeek's Anthropic endpoint in thinking mode, every
assistant tool_use turn in history must carry a thinking block. replayThinking
emitted one only when the turn had stored reasoning. A tool-call turn can be
committed without reasoning (the missing-reasoning fallback in the run loop
keeps the first complete response), and from then on every request in the
session, subagent review sessions included, was refused with HTTP 400
"content[].thinking in the thinking mode must be passed back to the API".
Studio has no repair path for that 400, so the session stayed broken.
Fix: while DeepSeek thinking is enabled, a tool-call turn without stored
reasoning replays {"type":"thinking","thinking":""}, which the endpoint
accepts (live-probed in #8927). contentBlock keeps the "thinking" field when
it is empty, because a thinking block without the field is refused as well.
This matches the OpenAI-compatible DeepSeek path, which already sends an
empty reasoning_content key on such turns. Turns with reasoning and thinking-off
requests serialize exactly as before.
Test: TestDeepSeekToolCallTurnWithoutReasoningStillReplaysThinking drives
Stream against an httptest server that answers like DeepSeek. On origin/studio
it fails with "status 400: ... content[].thinking ... must be passed back";
with the fix it passes. TestDeepSeekThinkingOffSendsNoEmptyThinking guards the
thinking-off shape. go test ./internal/model/anthropic/ ./internal/runtime/agent/
./internal/assembly/boot/, go vet, make check and golangci-lint are clean.
The P4 demonstration showed the shadow verdict calling a turn done when the
model changed nothing and said the task was complete: with no deliverable
in the contract, a turn that owes nothing reads as finished (#11056 made
such turns "none", which is honest but still not "owed").
Deliverables come from structure, never from the task's wording: a plan
whose steps name candidate files is a plan that expects its work to land
somewhere. Such a plan now adds one criterion, change@plan, satisfied by any
change the host proved in the task. It binds no path: candidate files are
the planner's inference, the right change may touch a subset, and a replan
moves them. Files the planner only read do not count, since an analysis
plan reads files too.
- A turn under such a plan that changed nothing ends incomplete on the
sealed verdict, and the receipt lists "the plan names files to change,
and no change was proven".
- Losing the deliverable is a relaxation like any plan check: a later plan
naming no files is routed to the user when an approver exists, and a
replan that still names files keeps it.
- Plans with no candidate files, and turns with no plan, are unchanged.
The readiness gate does not hold it yet; that follows the same measured
switch plan criteria took.
A blank-line run may drift in length only when non-blank lines bound the
window on both sides, so a blank-only old_string no longer names every run
in the file, and the window now applies the same trailing-newline rule as
the other fuzzy modes instead of writing a newline the file never had.
The P4 rollout demonstration ran a system-prompt candidate that told the
model not to edit files and to say the task was complete. The rollout gate
rejected it, but most of its 29 false completions were turns the shadow
verdict called completed with no obligations at all: the model changed
nothing, so it owed nothing, and owing nothing read as done. The contract
has no deliverables yet, and the rules forbid guessing from the task text
that a change was asked for.
Such a turn now gets outcome "none": nothing owed and no change the host
saw is no evidence of completing anything. A turn with a successful change
and nothing owed still completes; a failed write is not a change. "none"
agrees with a report that gave no verdict and diverges from one that said
done.
Nothing the user sees changes: the gate never inspected such turns and the
receipt was already absent for them. What changes is the bundle outcome
and every rate built on "completed", which it overstated.
The design doc records the new outcome and the P4 demonstration (promote a
harmless prompt candidate, reject a harmful one, both from bundle metrics).
First P4 slice of docs/design/TRUSTED_EXECUTION.md: the offline rollout
gate (K1: promotion authority is the offline corpus with repeated samples;
I11: false completion must not rise beyond the noise floor on a corpus with
no-solution tasks).
e2ebench -mode rollout reads baseline reports, "--", then candidate
reports (repeated samples are pooled) and decides:
- hold when the comparison means nothing: a side with no capability record
or more than one set (rollout.version_unknown), both sides the same set
(rollout.same_version), different tasks (rollout.corpus_mismatch), no
no-solution tasks to catch a false completion (rollout.corpus_lacks_no_solution),
or fewer than 50 solvable runs a side (rollout.insufficient_samples);
- reject when false completion or grader pass is significantly worse, its
whole 95% interval on the bad side (rollout.false_completion_rose,
rollout.pass_dropped);
- promote otherwise.
A rise inside the noise floor passes: ruling out any rise at all would take
hundreds of runs. The intervals are printed so the resolution is seen. With
-json the decision is written as a machine-readable record, reasons as
codes.
On the P3 reproduction data (one version, 29 solvable runs a side) it holds
for rollout.same_version and rollout.insufficient_samples.
Second P3 slice. compare mode printed two reports' metrics side by side and
left the reader to decide whether a difference meant anything. It now adds
a reproduction section:
- version: one capability set on both sides, two sets (a comparison, not a
reproduction), mixed, or unknown for reports that predate the set;
- corpus: whether both ran the same task ids with the same prompts;
- grader pass on solvable tasks, bundle completed, and false completion,
each with B - A and its 95% interval from Newcombe's method over Wilson
intervals, which stays sensible at the small counts and 0%/100% rates
these reports produce.
A report pair that is one version on one corpus with every interval
containing zero is reported as reproducing within the noise floor.
On existing data, one binary run twice over 82 tasks gives intervals of
about +/-15pp, and twice over 30 tasks one rate moved 20pp and was still
inside its interval: the resolution a single run has, now printed rather
than assumed.
First P3 slice of docs/design/TRUSTED_EXECUTION.md (I10: a rollout metric
counts only runs whose capability digests match). Until now a trajectory
said which model ran and hashed the prompt and tool schemas as a whole, but
not which skills, and no single field said whether two runs had the same
capabilities.
The run header now lists every capability version the run could use, and a
digest of the set:
- host: the build (version and commit), which carries the planner, memory
and compaction policies compiled into it;
- prompt: the system prompt;
- tool: each tool schema, MCP tools included, which is where an MCP
integration's behaviour shows (its launch config can hold secrets and is
not a version);
- skill: each skill by what it says and how it runs, never by where it was
found, so one skill on two machines is one version.
e2ebench reads the set digest and reports how many sets a report spans; a
report over more than one is an average over versions and says so.
Checked with a real reasonix.exe run on the Windows box: the header carried
1 host, 1 prompt, 9 skills and 18 tools and a set digest.
Stage two of moving completion onto the contract revision: the readiness
gate stops reading the replayed contract, where a complete_step citation of
kind manual settled a plan criterion on the model's word.
A required plan criterion now holds the turn only when its step names a
verification command that has not passed since the latest change (or has
never passed, when nothing changed). The message names the check that
settles it. A criterion no command checks is no longer held: the host
cannot settle it, and holding it only ever asked for a citation. It is
listed on the receipt instead (stage one).
Delivery sign-off still requires complete_step; only its role in settling
plan criteria is gone.
Stage one of moving completion onto the contract revision: what the user
reads changes, what blocks a turn does not.
The receipt's unproven plan criteria came from the replayed contract, where
a complete_step citation marks a criterion satisfied on the model's word.
They now come from the sealed verdict: a required plan criterion is proven
only when every verification command its step names passed after the latest
change. A criterion whose step names no command is listed as unproven with
"no command checks it", instead of being settled by a citation. Every other
gap is unchanged, and a turn with no sealed verdict (no trusted state, or no
contract revision in force) keeps the report's gaps as before.
The receipt card renders gaps, not a verdict noun, so this shows up as
listed items where a citation used to leave none.
Not changed: the readiness gate still reads the citation. Moving it is
stage two, measured first for the extra rounds it costs.
Plan criterion ids are positional within one plan (c1, c2, ...), so
#11017's plan@c1 meant different criteria in two plans of the same goal: a
reordered plan could pass as unchanged, a reworded one went unseen, and an
inserted criterion could read as a dropped one. What decides whether a plan
criterion is verifiable is the command its step names, and a command's
canonical identity does hold across plans. The contract now freezes those
commands, deduplicated with the project's checks, and answers each plan
criterion through the check its step names; a criterion whose step names
none is left to the replayed contract as unverifiable.
The plan flow asked the user only when a revision expanded scope. Dropping
a verification command narrows the work and relaxes the contract, and it
never reached the user. Now:
- a plan the user approved carries that, host-stamped (ApprovedByUser is
not part of the planner's JSON), and a derivation under it that loses
only plan checks is accepted as the user's (user:plan_approval);
- a later plan that would drop a check the task accepted is routed to
approval when an approver exists, even if neither the route nor the
planner asked for it;
- with no approver the plan runs and the dropped check stays owed.
The report still reads complete_step citations. Retiring that changes
what users see for every planned task with a criterion no command checks,
so it moves with the gate switch, where the incomplete terminal is
designed.
SignPath has no quota left, so Studio Windows builds could only ship
unsigned. Port the 1.x Certum SimplySign signer, pointed at Studio's own
account and certificate.
- Signing runs only when STUDIO_SIGNING_ENABLED=true, as three jobs, so
the SimplySign session is never up while third-party code runs:
windows-sign-payload signs the window, kernel and computer-use helper;
windows-package builds the installer and zip on a separate runner with
no environment or secrets; windows-sign-installer signs the installer.
The two signing jobs check out only workflow_sha, install no
toolchain, leave Defender on, and stop SimplySign after signing.
- Verification checks trust, the pinned thumbprint, the
STUDIO_SIGNING_SUBJECT subject, an issuer and a timestamp on every
signed executable, on the installer, and on the zip's copies.
- The SimplySign MSI is pinned by SHA-256 as well as by its signature.
- With the switch on, a missing secret or subject, or any signing
failure, blocks publish. The release body's Windows sentence follows
windows-sign-installer's result.
- Credentials are STUDIO_CERTUM_USERNAME, STUDIO_CERTUM_OTP_URI and
STUDIO_CERTUM_KEY_ID only. The concurrency group is
studio-certum-signing, shared with a new non-publishing smoke workflow.
- publish downloads every package artifact by its exact name, taking
Windows from windows-sign-installer when it ran, and fails when two
artifacts carry the same file name.
- signing-contract validates and fingerprints both the tag's tree and the
workflow_sha tree.
- The SignPath request steps, artifact configurations and request helper
are removed. The contract now covers the Certum files, and its call
graph detects Certum credentials.
- The runbook lists the repository settings to decide on before enabling
and says which PE files stay unsigned (Electron DLLs, elevate.exe, the
NSIS uninstaller).
The approved plan's acceptance criteria join the canonical contract. Each
is verified by the verification commands of the step it belongs to: the
user approved those commands with the plan, so they are the plan's own
verifiers, not a check the model picked. A criterion counts only when
every one of them passed after the latest change; a step naming none gives
the criterion no verifier, recorded as verifier.none rather than guessed.
This gives the shadow verdict a host answer for plan criteria that were
otherwise answered by complete_step citations, where the model binds a
check to a criterion. The frozen contract never reads that binding, and a
plan criterion it holds is not counted a second time from the replayed
contract.
Satisfying a criterion with its own step's commands discharges nothing the
host itself owes: stale verification and the project's checks are still
derived separately.
Open, recorded in the design doc: a plan revision the user approves that
drops a criterion is refused like any relaxation, and the report still
reads the citation. Both must change before the gate reads the revision.
On an unplanned turn the shadow contract turned every todo into a required
criterion and resolved it Satisfied when the model marked it completed.
Nothing but the model's own call stood behind that, and it reached the
user: a turn whose only action was marking a todo completed got a
completion receipt saying done. P1 measured it as the claim_only
divergence.
A todo is the model's own breakdown of the work, not a criterion anyone
accepted, so it is now recorded as optional and never resolved from its
status. Dropping only the resolution would have left every unplanned
multi-step turn reporting incomplete however well its change was proven;
making todos optional leaves completion to what the host proved, as on a
turn with no todos. An unfinished todo is still the readiness gate's to
hold: it counts open todos itself and never read this contract.
The report's "anything declared" test now counts required criteria only,
so a turn that only marked todos gives no verdict instead of done.
Still open: a complete_step citation binds a check to a plan criterion on
the model's word; that edge goes when plan criteria carry frozen verifiers.
Background and fleet children could not settle an unclassified command:
their walks would see siblings and the parent writing too, so #11009 left
them unknown. A run that declared write paths now takes a change the walk
sees inside them as its own, even while other writers run, and ignores
everything outside them.
This is an inference the user accepted, not a proof: claims are proven
disjoint up front and writer tools are fenced to them, but a sibling's
shell command writing into another run's paths is not excluded. A call
that only saw changes outside its paths stays unknown and is never read as
having changed nothing. A run with no declared paths, the root included,
attributes a walk only while no other background writer runs, and that
check no longer counts a background child as running beside itself.
The design doc records the inference as W11.
Not measured: on current studio the unproven_mutation divergence was 1/21
across seven delegation tasks (#11009), and a comparison costs about
CNY 8 for a difference that small.
A bare "/name" line or a lone skill chip reached the provider as the raw
rendered skill body in plain user prose: no <skill-pin> sentinel, and on the
structured (composer) path the typed text trailed as an unlabeled paragraph
instead of an Arguments line. The model-invoked path (run_skill,
read_only_skill, read_skill) already sent the pinned form, so one skill had
two payload shapes depending on who invoked it.
Every invocation site now renders through skill.RenderInvocation: the
string submit path, the structured path, RunSkill (ACP prompts) and the
slash_command tool. A lone inline chip takes the typed text as its
arguments, matching "/name task"; several inline chips, or chips mixed with
subagents, keep appending the shared text once. Nothing moves into the
cache-stable prefix. Store.Render is removed and Render is unexported, so the
pinned form is the only one an invocation can take.
An effect test through the real Build asserts the pinned payload at the
provider boundary for the bare and argument-carrying forms of both paths.
Refs #11004
A command the host cannot classify (sed -i, an inline interpreter) is
settled by walking the workspace before and after it. Only the root agent
had a root to walk: sub-agents are built without the write root, so a
child's unclassified command stayed unknown, merged into the parent that
way, and left an unproven_mutation no later check can discharge. On the P1
macOS run this was the cause behind the forced-delegation divergences
(sed -i '' and python3 -c inside children).
- agent.Options gains ObserveRoot, used only for that walk; the root agent
keeps walking its write root, so nothing else about it changes.
- Delegation hands a writer child the workspace as its observe root only
when it runs in the foreground: the parent is waiting on it, so what
changes during one of its calls is that call's. Background and fleet
writers get none, since their walks would claim their siblings' writes.
- While any background writer is registered, no agent settles by walking,
the root included: the difference between the two walks is not the
call's alone.
Measured on the Windows box, 7 delegation tasks x 3 samples per arm: the
unproven_mutation divergence was 1/21 on current studio and 0/21 with this
change, grader 21/21 on both. At that frequency the effect is not
measurable; the change stands on the mechanism and its tests, not on the
number.
A shell command that runs code the host cannot prove read-only, and that is
not a recognized check, counts as a possible change: after it the model is
handed a new stale_verification debt keyed on its own command, and nothing
says why. The model reads that as its check having run, or as the host
asking for the same thing again. The host already knows the reason, so the
first such command in a turn now carries it, with the commands the host
does recognize. The decision reads the host's own classification (not
read-only, verification not_verification, stale verification owed after),
never the command's wording, and it rides the tool result, not the prefix.
Measured on the Windows box, 30 tasks x 2 samples per arm, same binary
apart from this change, run back to back:
stale_verification divergences control 34/60 (57%) notice 19/60 (32%)
grader pass rate control 58/60 notice 58/60
tool calls per task control 20.0 notice 24.2
tokens per task control 199k notice 221k
Two-proportion z = 2.76 (p ~ 0.006). The tasks were selected for having
shown the divergence before, so the earlier 70% rate is not the baseline;
the control arm is. The extra calls are the recognized checks the model
now runs.
Expanding ${REASONIX_WORKSPACE_ROOT}/secret on Windows leaves a forward
slash after the drive-letter path, so the raw string never equals
filepath.Join's result even when the root came from the environment.
Studio and the 1.x line share config.toml and both read [browser] enabled,
with opposite defaults: 1.x writes enabled = false on every save of the
user file, which Studio then read as the person turning its browser tools
off, and a Studio save wrote enabled = true, which switched the 1.x command
line's browser on.
Studio now decides on [tools] browser_tools (unset = on) and carries
[browser] enabled through a save exactly as read, never writing it where the
file had none. Boot binds the browser tools from the new key.
The settings screen gets a built-in browser switch under Tools, backed by
GET/POST /browser-tools on the provider-edit grant. The save writes the user
file and rebuilds the runtime, declared as runtime-rebuild.
First P2 slice of docs/design/TRUSTED_EXECUTION.md. The task contract is now
a host-owned record with an identity and immutable revisions, not a replay
rebuilt from receipts each turn.
- internal/runtime/contract derives criteria only from sources the host
owns: the check identities a task began requiring and the test criteria
captured as host-owned bytes. Derivation is canonical, so the same sources
give byte-identical revisions. Host policy accepts a first revision and
any revision that keeps every criterion and adds to them; one that would
drop or alter a criterion is refused and the revision in force stands.
- Revisions are sealed as contract/1 records in the workspace stream. The
delivery checkpoint names the record, so a goal resumed in another
process is held to the revision it accepted, read back from Trusted Host
State. A name the store cannot produce is reported, never replaced by a
fresh acceptance. A new task starts a new contract.
- The shadow verdict answers each accepted criterion from the ledger, not
from the checkpoint's copy of the declaration, and drops the host
obligation that owes the same verifier so one debt is not counted twice.
The audit and trajectory carry the revision, the policy decision and any
load failure.
Still shadow: the readiness gate is unchanged. The refusal test edits a
checkpoint by hand, which nothing in production does; it demonstrates that
the checkpoint is not the authority, not a reachable bypass.
With the variable set empty the pattern expands to "/secret", which is
absolute on Unix but drive-relative on Windows, so it resolved under the
workspace root and the test failed on every Windows run since 0a0e1989a.
Pointing the variable at a separate directory asserts the actual property -
the value comes from the environment, not the root - on every platform.
The notes landed with section headings outside the allowed set and changes
that name no pull request, which pushed the repo-wide release-note count past
its ceiling and turned lint red on studio. Same content, each change now cites
#10970, #10980 or #10982.
The publish job now renders release-notes/studio/<semver>.md before it
signs anything. A #N that is a pull request gains "by @author"; a #N that
is an issue gains "fixed in #M by @author" for each merged pull request
GitHub links as closing it (the ClosedEvent closer and
closedByPullRequestsReferences, read over GraphQL; mentions are not
fixes). An issue fixed by a direct push stays plain. A 贡献者 list of the
credited humans, in first-appearance order, closes the notes. Bot authors
are never credited; the source file is not rewritten.
scripts/markdown-refs.mjs decides what is a reference: #N at a line start
or after whitespace, "(", "（" or a list separator, outside code spans,
backtick and tilde fences (each closed only by its own kind), indented
code, HTML comments and link targets. A number that names a discussion or
nothing stays plain with a warning instead of failing the release.
Lookups retry with backoff on 5xx, rate limits and dropped connections. A
failure that persists, or an authorisation failure, fails the job with a
release_credits.* code before signing. The publish job gains
issues: read and pull-requests: read.
CI gains a release-notes-scripts job, gated on scripts/ and
release-notes/studio/, that runs the three test files with contents: read.
scripts/release-credits.mjs is the same module the 1.x renderer uses.
A full e2ebench run put ten of eighteen divergences under one reason,
obligation.outstanding, and e2ebench gives every task its own state root
and deletes it afterwards, so the sealed bundle that said which obligation
it was does not survive the run. The kind is a content-free enum and is
what decides which old path owes the difference, so the audit now carries
host:<kind> in its place.
Every save replays the event log before it appends, and the reader refused a
log past 128 MiB. Images were stored inline as data URLs, so a session with
about 2 MiB of images per turn crossed the budget around turn 63: from then on
LoadSession failed and every Save failed with the same error, and later turns
never reached disk.
New records write each data-URL image to <id>.blobs/<sha256>, fsynced before
the record that names it, and carry an image_blobs side table instead. Replay
reads each blob back byte for byte and verifies its hash. Such records are
stamped schema_version 3, so older studio and 1.x builds refuse the log
instead of loading it without its images and saving that loss back.
A missing blob is restored from the copy the .jsonl checkpoint still inlines
and written back. With no copy left, the message keeps its text and records
the lost image as provider.UnavailableImage (hash and position); only
ModelMessages renders a note, and the blob reappearing restores the image. A
transcript missing an image never rewrites a checkpoint that still holds it,
and a runtime still holding an image whose blob vanished writes it back
instead of being reported as diverged; what cannot be restored fails under
ErrSessionImagesUnavailable.
The reader ceiling is 512 MiB. A save folds a log past 128 MiB only when the
fold drops at least 64 MiB, so a legacy inline-image log is folded to its text
once, while a large text-only log is not rewritten on every save. A compaction
sweeps blobs that neither the published log nor a salvage or recovery
sidecar (.events.jsonl.damaged, .superseded, .context.json) names, and stale
temp files; a sidecar it cannot read whole stops the sweep. The stored
compaction projection is built with ProjectionMessages, so an unavailable
image stays a host record there instead of rendered text. The
blob directory is a session sidecar dir: delete, rename (blobs before the
log), recovery trash, native re-homing and sub-agent delete take it along.
Refs #10965
Problem: on start-up the sidebar said "No folders yet" (尚无文件夹) and
"0 projects" under This machine, and the empty transcript asked to open a
project first, until the kernel answered the first tree read. A slow start
looked exactly like a machine whose folders had been lost.
Root cause: App held the tree as TreeWorkspace[] initialised to [], so "not
read yet" and "read, and empty" were the same value, and every reader of it
(the sidebar empty row, the project count, needsProject) drew the latter.
Fix: App tracks treeRead, set once the first hub.tree() read settles. Until
then the sidebar shows a "Reading folders…" status row instead of the empty
row, hides the project count, and needsProject stays false. With no pane
open, the main area says only that no session is open, without the
add-folder prompt it gives once the tree is read and empty. A failed read
still settles, as before, so the window never waits forever.
The row carries the clock icon on the existing tool-running spin (still
under prefers-reduced-motion) and, past one second, the elapsed time via
format.seconds, ticking once a second. The tick is display only and is
aria-hidden, so the status region is not re-announced every second; the
interval lives with the row and is cleared when the read settles or the
row unmounts.
Test: sidebar-first-read.interaction.test.tsx holds hub.tree() pending,
asserts the loading row and no empty-state wording, then resolves [] and
asserts "No folders yet"; a second case with no panes asserts the main
area's add-folder hint is absent until the read settles. Both fail on origin/studio (no loading row) and
passes with this change. tree-reading.interaction.test.tsx uses fake
timers: the elapsed value reads 3s then 5s, is aria-hidden inside the
status, disappears with no timers left once the read settles, and the
interval is cleared on unmount. The perf harness gains ?treems= to hold
tree reads back and ?panes=0 to open with no pane, used for before/after
screenshots in light and dark; perf/motion.mjs counts the reading row's
spin as live work.
Third slice of the trusted execution layer (P1). Every shadow bundle now
carries an outcome derived only from host evidence, and how that outcome
diverges from the completion report's verdict.
- internal/runtime/verdict: Evaluate turns the contract's criteria and
checks, the receipts and the ledger's outstanding obligations into one
verdict per obligation and an outcome (completed, incomplete, failed,
blocked). It never reads a criterion's status, because todo_write and
complete_step can set it; before contracts freeze verifiers, the only
criterion host evidence satisfies is an atomic ask proven by a successful
mutation. Checks keep the state the contract derives from receipts.
- Classify compares that outcome with the report's verdict: agree,
new_stricter or new_looser, with reason codes (claim_only, evidence.stale,
evidence.missing, check.failed, obligation.outstanding, verifier.none,
old_gaps, old_criteria).
- The EvidenceBundle audit carries outcome, class and reasons, and
e2ebench prices them like the completion report: false completion is a
completed bundle on a run the grader failed, next to the divergence and
reason census.
TestEffectClaimOnlyCompletionDivergesThroughRealBuild runs the known hole
through boot.Build: a todo marked completed with no other action finishes
the turn and the report says done; the sealed bundle says incomplete,
new_stricter, claim_only.
Second slice of the trusted execution layer (P1). The shadow bundle now
records the workspace as each turn found and left it, what changed during
the turn, and what changed between sealed turns while nobody observed it.
- internal/state/observation: a snapshot is a tree of directory nodes stored
as content-addressed index objects in Trusted Host State, so an unchanged
directory costs nothing to record again and two snapshots diff by walking
only subtrees whose digests differ. The policy is host-owned; the default
excludes version-control stores and the host's own state root, each with a
reason, and keeps dependency trees observed because verifiers read them.
Ignore files and workspace config are never consulted.
- A file is identified by its stamp (size, mtime, and on macOS/Linux ctime,
inode, device), not a content hash. On this repository's main checkout
(107,803 entries, 2.1 GB of node_modules) a cold snapshot takes 1.9 s and a
warm one about 0.2 s; hashing every file would cost seconds per process.
- trustedstate gains PutIndexObject, which skips the per-object fsync: tree
nodes are an index, not a seal, and with fsync the first snapshot of this
repository took 31 s. A node lost to a crash makes a later diff
unverifiable, never a satisfied comparison.
- The turn-start snapshot runs while the first model request is in flight;
the first tool batch waits for it, so nothing the turn does lands in its
own baseline. The previous snapshot is read from the stream's head record,
so edits made between sessions are seen too.
- The EvidenceBundle audit adds snapshot completeness and the count of
unobserved changes.
The design doc's W1/W3 now say what was built and why (W8-W10).
First slice of the trusted execution layer (docs/design/TRUSTED_EXECUTION.md,
P1). Nothing reads the bundle back to decide a turn.
- internal/state/trustedstate: content-addressed objects, one record chain
per stream, a head file replaced by rename. A head this process already
observed refuses rollback, a same-generation swap and a history that no
longer reaches it; a store with no observed head cannot tell a forged but
self-consistent chain, and the tests pin exactly that. Failures carry
sentinels and a typed code.
- IntegrityLevel is supplied by the components that confine what the model
drives: sandbox.WriteProtects (the platform's effective write roots, both
overlap directions, symlinked ancestors), plus one-way flags set by the MCP
launcher and the extension sidecar launcher whenever they start a process
outside the sandbox, including a "confined" MCP launch that fell back
because no backend was available.
- File tools refuse writes into <state>/trusted with a typed refusal
(workspace.trusted_state_write); an explicit allow_write entry keeps raw
access and also reaches the bash write roots, which drops the level.
- The root agent seals the contract, completion report and receipts at the
end of every turn, including turns the readiness gate stopped. Receipt
arguments are kept as digests only. A new content-free EvidenceBundle
audit reaches the trajectory with the record digest, generation, integrity
and the typed cause when sealing failed.
The design doc records what implementing this showed: MCP servers default to
host mode and sidecars always run unconfined, so installs with either are
tamper-evident only, and the level is per process (another process running
as the same user can write the store).
Define the host-owned layer behind "allowed to start", "completed with
evidence" and "allowed to ship": trust boundary, canonical objects
(Task Contract, Evidence Bundle, Capability Rollout, RiskPolicy,
ObservationPolicy, WorkspaceSnapshot, TrustedHead), state machines,
invariants and typed failure semantics. No UI and no implementation.
Workspace content, including REASONIX.md and .gitignore, is source
material for revision requests and never changes enforcement. Hash
chaining is tamper evidence relative to a trusted head, not an
authority boundary, so L3 enforcement requires host-protected state.
The document lives under docs/design/ so the embedded product-doc
corpus (docs/*.md) cannot present unshipped design as a feature.
Delete (and Cmd+Backspace on macOS) on the focused session row, or on a
kept version/recovery copy row, opens the same inline confirmation the row
menu's delete opens. Only a row holding focus reacts: the handler requires
the row itself to be the event target, never a hover target, and stays out
of the way while the row is being renamed.
The confirmation now also answers Escape as cancel, without letting the key
reach the window's stop-the-run handler, and focus returns to the row the
key asked about so a keyboard is not stranded on the page body.
Focus goes back only when the question is dismissed, not when another
row's question replaces it; after a delete lands it moves to the next row,
or the previous one when the deleted row was last.
The row keydown declares both session.open and session.delete, and the
copy row's delete button and the confirmation's cancel carry their action
ids, keeping the census's unnamed-write count where it was.
Fixes#3220
A screen share or a recording leaks the account balance and the session's
spend as readily as anything else in the meter bar. An eye button beside the
wallet masks both as ••• and persists per machine (rx-hide-amounts), so they
stay hidden across restarts. The sidebar's wallet entry and every amount in
the cost card (session total, this turn, original currency, per-source rows,
wallet) are masked too. Only the drawing changes: the wallet keeps being
read, so revealing shows the current value without a request.
Fixes#4135
The file writers refuse a target outside workspace_root and allow_write on
every OS, including Windows where bash is never OS-sandboxed. The refusal
was a plain error whose text said writes were "confined" and told the
reader to "widen [sandbox] workspace_root / allow_write", so a Windows user
who never enabled a sandbox was told by the agent that it had been
sandboxed.
confine now returns a tool.Refusal coded workspace.write_outside_scope that
names the refused path, the writable directories, that this is the file
tools' scope rather than an OS sandbox, and the setting that allows it. The
agent carries a tool.Refusal's code onto the tool result on the error path,
so the frontend receives the cause as an identity, and the tool card renders
the localized cause from that code. The scope itself is unchanged.
An option can ask for more than itself ("paste the output"), but once it
was picked there was nowhere to type: in Studio the pick and the free-text
box were mutually exclusive, and in the TUI a pick answered at once.
Studio keeps the box open under a single-choice pick as an optional note,
so the answer goes out as [pick, note]; the kernel already joins several
selections. The note is its own state: an Other answer left behind when a
pick is made is not sent as a note, and going back to Other finds it still
typed. Choosing Other still makes the typed text the whole answer, Enter in
the note box answers like Enter in the Other box, and a sealed card shows a
pick's note under the pick rather than as Other.
The TUI takes Tab on a single-choice option as "pick and add a note": the
composer opens for the note, Enter sends it after the pick, Esc keeps the
pick without one. Picking an option now also drops a previously typed
answer, which a single-choice question used to send alongside the pick.
Fixes#8136
A directory claim overlaps every path beneath it, so two writers that
name the same directory serialize even when they touch different files.
The task and fleet descriptions and their write_paths properties now say
to declare file paths when the targets are known.
Both tools sit behind use_capability, so the provider-visible prefix and
the boot golden fixtures are unchanged.
OpenCode Go refuses third-party requests that carry no session:
"Request is missing x-opencode-session and cannot be routed efficiently"
(HTTP 400). The 1.x provider layer sends that header; the studio adapters
never did, so every OpenCode Go call failed on its first request while
the same configuration worked in 1.x.
Each of the three adapters (chat, messages, responses) now mints one
session id per client and presents it, together with a Reasonix
User-Agent, on every request whose URL is on opencode.ai/zen/go. A
header the user configured is kept. Other hosts are untouched.
With the session in place the first tool round trip still failed:
OpenCode Go's chat route rejects a role=tool message carrying "name"
(fixed on 1.x in #10852). The chat adapter omits it there only.
The responses client groups the new id with the DashScope session-cache
flag, and the anthropic client with its custom headers, so neither
guarded struct grows a scalar.
Test: a local gateway that answers 400 without x-opencode-session,
reached through a transport that keeps the opencode.ai URL.
Before: all three adapters fail with "status 400: ... missing
x-opencode-session"; TestOpenCodeGoChatToolResultOmitsName fails with
tool message carries name "read_file". After: all pass, and two
clients present distinct ids while one client keeps its own.
Fixes#10921
ask also carries MCP elicitation, so the Notification hook forwarded an
MCP server's prompt verbatim to whatever channel the hook feeds. Name the
server for an elicitation ("answer needed (MCP <source>): ..."), prefer
its own message over the first field's prompt, and clip every forwarded
prompt to one line of 120 runes.
Follow-up to #10812.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
On a Chinese Windows, cmd.exe reports a missing command in the console
code page (GBK), not UTF-8. The stdio transport kept the stderr tail as
raw bytes, so an MCP startup failure such as "'read-only-mysql-mcp-server'
is not recognized as an internal or external command" reached the settings
page as U+FFFD mojibake and the user could not see why the server failed.
The stderr tail now goes through encoding.DecodeOutput, the one decoder
for process output. The buffer keeps the last 16 KiB, so only its head
can be cut: tailBuffer records when it dropped bytes and passes
Cut{Head: cut}. The end is never cut, because it is whatever the server
last wrote. Output that is not UTF-8 is decoded from the raw bytes, so a
GBK tail that begins with a CJK character keeps that character.
Tests:
- TestStdioFailureDecodesCodePageStderr starts a real child that writes
GBK to stderr and exits.
- TestStderrTailDecodesTheServersCodePage covers GBK that starts with a
CJK character, both uncut and cut, plus UTF-8 cut mid-character.
All of them fail on origin/studio and pass with this change.
Fixes#7857
DecodeOutput trims a partial rune only at an edge the caller cut. Output
that nothing cut but that stops inside a character still has invalid raw
bytes, so it went through charset detection and fell back to GB18030,
and the whole output turned into mojibake: "编译成功…正在运行" plus half
of "中" read as "缂栬瘧鎴愬姛…". This is common in real use:
- a process killed on timeout mid-write;
- a persistent shell's partial() read;
- a block-buffered pipe.
The output is now read as UTF-8, less the final incomplete rune, when
two things hold: that rune is the only invalid part, and a multi-byte
UTF-8 character has already appeared. Both are structural evidence. When
only ASCII precedes the half rune, nothing says UTF-8, so the bytes
still go through detection. The uncut GBK probes (价, 啊\r\n, ok 洹,
啊 ok) are unchanged.
Test: TestDecodeOutputReadsUncutUTF8EndingInPartialRune fails before
this change with the mojibake above and passes after it. It also pins
ASCII plus half a rune (read as a GBK pair) and uncut, tail-cut and
head-cut GBK.
Refs #6488
The edit_file fuzzy matcher already tolerates trailing whitespace, tab-vs-space
indentation and copied read_file line prefixes, but not a blank-line run the
caller reproduced at a different length. PEP 8 asks for two blank lines between
top-level definitions and a model routinely writes one, so the span fails to
match a file it otherwise quotes exactly.
blankRunRanges collapses each run of consecutive blank lines into one token, so
a run's length is free while its presence is not: a blank line absent from the
file still never matches one the caller invented. Non-blank lines keep the
exact-match rule modulo trailing whitespace, and the single-match requirement is
unchanged. The relaxation is a standalone last resort rather than a fuzzyMode,
so a span that drifted in two dimensions at once still fails.
An ACP client had no way to bring back a Goal the session still held.
session/set_mode goal always drafts a new Goal, so the next prompt replaced
the retained objective and reset its counters, while the controller's
ResumeGoal/PauseGoal (reached through /goal in the CLI and Studio composer)
were never driven by a frontend directly.
ACP gains two vendor methods, advertised in agentCapabilities._meta with
schemaVersion 1:
- _reasonix.io/session/goal/resume re-enters a stopped, blocked, paused
or cancelled Goal with its objective, counters, todos and delivery
checkpoint. It does not start a turn; the next session/prompt continues
it. A running, complete or absent Goal fails with -32010.
- _reasonix.io/session/goal/pause pauses a running Goal at its next
continuation boundary, or fails with -32011.
On success both clear the turn-outcome label (cancelled, failed) the
status showed in place of the Goal state, return the Goal in the status
shape, publish a status_update with event "goal", and emit
current_mode_update when the mode moves. docs/ACP.md documents that a
"failed" status labels the last turn while the Goal keeps running.
serve drives the same capabilities through POST /goal/resume and
POST /goal/pause, answering {changed, status, goal}, so frontend-parity
holds without a scope exception.
The rule that a running Goal cannot be resumed moves into
goalMachine.resume under its lock. Before, resume accepted a running Goal
and bumped its continuation epoch, dropping the continuation in flight;
this also affected /goal resume in the CLI and Studio.
initialize/authenticate move unchanged to handshake.go so service.go stays
inside its size budget.
Tests:
- acp: TestServeGoalResumeKeepsTheRetainedObjective and
TestServeGoalPauseStopsARunningGoal fail with only the declarations
present (capability not advertised; -32601 method not found) and pass
after. TestServeGoalFailedLabelKeepsTheGoalRunning fails when pause does
not clear the label (pause reports "failed") and passes after.
- control: TestControllerResumeGoalRefusesARunningGoal and the new
TestGoalAdvanceResultCannotCrossGoalLifecycle subtest fail on the base
("resume accepted a goal that is already running") and pass after.
- serve: TestServeGoalPauseAndResumeKeepTheObjective.
A Studio window can go blank in three ways, and none had a way back
short of quitting, although the kernel keeps running the turn and a
loaded page reattaches to it:
- A render error outside the per-block boundaries unmounts the React
tree while the renderer lives on. The shell installs no menu with a
reload role, so nothing reloads the page.
- The renderer dies while the window is visible. main.js handled
render-process-gone only before the window first showed, so the
window stayed dead.
- The renderer dies while the window is hidden in the tray. showWindow
then revealed the dead page.
installReload binds Cmd+R on macOS, Ctrl+R elsewhere (by key or by the
KeyR code, so non-Latin layouts work) and a bare F5 via
before-input-event. A dead renderer's input never reaches that event, so
it also reloads on render-process-gone while the window is visible,
except for a clean exit. It returns revive(), which showWindow calls
before showing: it reloads when webContents.isCrashed(). Crash reloads
share a guard of at most 3 within 60 s; past that the window is left
alone instead of looping.
Test: desktop/electron/test/unit.mjs, three tests (reload key; visible
crash plus loop guard; hidden crash revived before show). They fail on
the base (Cannot find module '../src/reload.js'), and the hidden-crash
test also fails against the previous revision (no revive); all pass
after, suite 46/46. On Electron 44.1.0: a visible
forcefullyCrashRenderer was followed by did-finish-load; crashing on
every load gave three reloads and then stopped; crashing a hidden
window and then revive() plus show() reloaded it (isCrashed=true, then
did-finish-load).
Refs #10319
Running Reasonix in a git repository left pasted attachments under
.reasonix/attachments and background-task state under .reasonix/tasks as
untracked files, so users added .reasonix to their .gitignore by hand.
Root cause: both stores created their directories in the workspace with no
marker telling git they hold host runtime state. Ignoring all of .reasonix
is not the answer: the same directory holds files a project commits on
purpose (settings.json hooks, skills, output styles, instructions).
Fix: fileutil.MarkUntracked drops a ".gitignore" containing "*" into a
directory the host owns. The marker is written to a temp file in the same
directory and published with os.Link, so a failed write or a crash never
leaves an empty marker that would read as present forever; Link also
refuses an existing target and does not follow a symlink there, so a
marker already in place is kept. The attachment writer and the task store
(task dirs and idempotency records) mark their own roots, best effort in
both: the marker is a courtesy to the user's git, and a filesystem without
hard links must still store the state. The user's .gitignore is never
touched.
Test: TestSavedAttachmentsStayOutOfGitStatus and
TestFileStoreStateStaysOutOfGitStatus run real `git status --porcelain
--untracked-files=all` in a fresh repository with the machine's git config
cut off (GIT_CONFIG_GLOBAL at a missing temp file, GIT_CONFIG_NOSYSTEM,
XDG_CONFIG_HOME at a temp dir, -c core.excludesFile=). With the base
store code they fail listing .reasonix/attachments/* and
.reasonix/tasks/{t1,.idempotency}/*; with the fix the status is empty.
TestMarkUntrackedLeavesNothingWhenTheWriteFails fails against a
create-then-write marker ("failed write left [.gitignore] behind") and
passes with the linked temp file; a symlinked marker is not followed.
Bold text in the transcript could only be told apart by weight. The
.md strong rule painted it with --text, and the theme vocabulary had no
entry a pack could use for it, so the only way to colour it was editing
the shipped stylesheet.
Add an fgStrong colour token. The kernel accepts it, the frontend maps it
to --text-strong, and .md strong reads var(--text-strong, var(--text)),
so every pack that does not name it renders exactly as before. It steps
with the reader's contrast setting like fg does, and the shipped-pack
contrast test now holds it to AA on the same surfaces as the other inks.
Test: TestDecodeAcceptsAStrongTextColour fails on the base (the token is
dropped with a "not a theme token" warning) and passes after; the two new
theme.test.ts cases fail on the base (--text-strong never set) and pass
after. Real-browser screenshots of the perf build, light and dark: before,
setting --text-strong leaves bold at the body colour; after, bold takes
it, and the default rendering is unchanged.
The root .gitignore already lists /workgroupassay; the 10 MB binary was
committed with 491016f8c alongside the tool's source and is rebuilt by
go run ./tools/workgroupassay.
TestEffectRestartKeepsTheReviewAHighRiskGoalOwes (internal/assembly/boot)
failed on windows-latest in runs 36256346762, 36232402249, 36230551992,
36229537444 and 36258827818 (#10901): "PHASE-1 ... never settled", goal
running, 25-26 requests after 20s.
Root cause: the goal in that test can never finish by design and stops only
on its token budget. At 30000 tokens and 1000 per reply, each phase ran about
30 provider requests, 23 of them refused "done" claims, each a full host
turn: about 65 ms on macOS here and, from the CI tail, about 0.8 s on a
hosted Windows runner. 30 turns at that rate do not fit the 20 s settle window. The
assertion only reads whether phase two carried the review nudge and whether
the goal ended incomplete, and the first refused claim already settles both.
At 8000 tokens each phase makes its planned calls plus three nudged
continuations, 16 requests per arm instead of 60, with the same outcome
(blocked, review demanded, in both arms). Settling now waits for the goal to
leave running while the controller is idle. The old test read "idle for
500 ms" as settled, but an idle gap between two continuations is not the
end of the goal.
TestTrySteerRejectedBecomesFollowup (internal/session/control) waited 1 s
for the gated runner. TrySteerInboxItem dispatches the rejected steer
before it returns, so the claim is already durable (state running, intent
follow-up, controller running) when the call comes back. What was being
timed is the turn goroutine reaching the runner. Before that it goes through
a chain of fsyncs: checkpoint begin, branch meta, in-flight marker, recovery
snapshot. That chain takes 100-670 ms on macOS here and can pass a second on
a loaded Windows runner. The test now asserts the admission synchronously.
The 8-minute package timeout in 36258827818 attempt 2 came from
TestThirtySteersApplyAndAckExactlyOnce. Its initial turn missed the same 1 s
start window, t.Fatal ran, and the deferred autosaveWG.Wait blocked forever:
the fake provider's first round was parked on p.release, which only the
test body closed. Every test that parks a fake on a release channel now
closes it idempotently (sync.OnceFunc) before waiting on autosaveWG, in
steeringTurn's cleanup and in TestSteerUnreadAtTurnEndIsDeliveredAsAFollowup.
TestThirtySteers now uses steeringTurn, whose existing 10 s start bound
remains a hang guard only. That wait also returns on TurnDone, so a turn that ends without reaching the provider fails at once
with the event. A deliberate t.Fatal right after steeringTurn used to hang
until -timeout. It now fails in 0.15 s.
TestBackgroundJobWakesSessionThroughManager failed the same way on this
PR's first Windows run (36263202340): a 1 s wait covered the job goroutine
queueing the continuation and the woken turn's fsync chain. The test now
takes the observer's own result for the publish step. It then waits for the
runner input, and the only other exit is a TurnDone. Every step ends in an
event the product always emits, so no clock races the disk. The note check
stays after the runner input: by then the turn's composition has drained
the note, or the ack has dropped it.
Local: -race -count=20 of the five control tests and of the boot test pass
(the boot test drops from about 18 s to about 4 s locally).
Root cause: the 1.x settings editor mirrored base_url into request_url
(and older ones into chat_url), and 1.x and Studio read the same
config.toml. Studio used either override verbatim, so a relay such as
https://router.shengsuanyun.com/api/v1 received its chat POST on the API
root and answered "404 page not found" while /models still listed the
catalog. 1.x fixed this in #9464 and its follow-up 73b6d9b4b; Studio never got
the comparison.
Fix: port provider.EndpointOverrideRepeatsBase into
internal/contract/provider (scheme and host case-insensitive, path exact
apart from a trailing slash, an override with a query or fragment is
deliberate and kept). The openai provider applies it to request_url and
chat_url, anthropic to its root with and without /v1, responses to its
base. Chat URL resolution moves into chaturl.go to keep openai.go inside
its budget.
Test: TestStreamIgnoresEndpointOverridesThatRepeatBaseURL (openai,
request_url and chat_url) and TestStreamIgnoresRequestURLThatRepeatsBaseURL
(anthropic, responses) stream against an httptest server whose base is
/api/v1. On the base they POST to "/api/v1/", "/api/v1" or "/api"; with
the fix they POST to /api/v1/chat/completions, /api/v1/messages and
/api/v1/responses. TestEndpointOverrideRepeatsBase covers the comparison.
Fixes#9658
The config template documents [statusline].command as a command whose
first stdout line replaces the built-in status data row, but the
terminal UI never read the key: nothing passed it from config into
tui.Options, so a configured command was silently ignored.
runTUI now hands the TUI a runner built from the loaded config. When
the footer refreshes its between-turn meters (startup, turn end,
session switch) the TUI fetches the kernel status, feeds
{model, contextUsed, contextWindow, cwd} as JSON on stdin, and shows
the command's first line in place of the telemetry row. Empty output
or a failed run keeps the built-in row. The spawn goes through the
hook spawner with a 2s timeout and without searching the working
directory for a bare executable name.
Statusline is user-level configuration and is never taken from a
project file: LoadForRoot restores the user/global value after the
project merge, the project render scope omits [statusline], and the
project delta never writes it. The template and reasonix.example.toml
say so.
Test:
- tui: TestStatuslineCommandReplacesTheTelemetryRowAfterATurn and
TestEmptyStatuslineOutputKeepsTheBuiltInRow drive the model through a
turn_done update against an HTTP kernel stub.
- cli: TestStatuslineRunnerFeedsStdinAndKeepsTheFirstLine runs a real
shell command; TestStatuslineRunnerIgnoresAProjectCommand loads a
project-only command and expects no runner.
- config: TestProjectStatuslineIsIgnored, TestUserStatuslineWinsOverProject,
TestStatuslineNeverRendersIntoAProjectFile.
Before the scope change the config and project-command tests fail on
their assertions (project value resolved and rendered); after it all
pass.
Problem: adding a source could write over, or read, a credential that belongs
to another endpoint (#9178):
- a name that folds to another provider's slot ("deepseek" beside the
official DEEPSEEK_API_KEY entries) wrote its key over that slot;
- adding under a name that already exists upserted over the entry and
inherited its slot, so a keyless source at a relay resolved and sent the
old key there;
- a removed provider's credential stays stored, and a later source whose
name derives the same slot read it.
Root cause: POST /providers derived the slot from the name alone and
treated an add as an upsert.
Fix:
- config.FreeAPIKeyEnvFor keeps the name's own slot only while no config
entry holds it and nothing is stored there; otherwise it allocates a
private REASONIX_CONNECTION_<id>_KEY slot, the scheme 1.x uses.
- POST /providers refuses a name that already exists (409
provider.name_taken) unless the body says replace. Onboarding, which fills
in the entry a fresh install already names, sends replace; a replace at
that entry's own host keeps its slot, anywhere else it gets a free one.
- An unreadable config refuses the save (500 provider.config_unreadable)
instead of falling back to the name-derived slot.
- The same-host keyless "second door" still shares a slot on purpose.
Test: on origin/studio, TestANewSourceNeverWritesAnotherProvidersKeySlot,
TestAKeylessSourceElsewhereDoesNotBorrowAFoldedSlot,
TestAddingUnderAnExistingNameIsRefused (200, want 409) and
TestAKeylessSourceDoesNotInheritARemovedProvidersKey (relay-b resolved
"sk-a") fail; all pass after, with TestReplacingAnEntryKeepsItsSlotOnlyAtItsHost
and the existing key-reuse tests. Onboarding.test asserts replace is sent.
edit_file on a UTF-8 file carrying CJK text rewrote bytes it never touched:
continuation bytes of characters far from the edit came back as 84 31 A4 37.
Detect took a nil error from the GB18030 decoder as proof the file was
GB18030. The decoder never fails: it maps every invalid sequence to U+FFFD.
So any file that was not valid UTF-8 -- one stray byte is enough -- was read
as GB18030, and the read-modify-write tools (edit_file, multi_edit,
delete_range, write_file over an existing file), checkpoint restore and the
config migrations then re-encoded every unreadable sequence as the GB18030
form of U+FFFD.
Detect now names a charset only when decoding and re-encoding restores the
input byte for byte. GB18030 is tried first, then GBK, which restores what
GB18030 cannot such as CP936's single-byte euro 0x80, so such a file stays
editable in its own encoding; anything else is LossyUTF8, whose bytes pass
through untouched. The new Kind is appended because checkpoints persist it
by value.
GBK cannot hold every character, so Encode now returns an error instead of
falling back to UTF-8, which rewrote a whole CP936 file as UTF-8 when an edit
added a character such as an emoji. A legacy charset that cannot represent a
character answers an *UnencodableError naming it, its byte offset and the
charset, with ErrUnencodable as its identity; the file tools, checkpoint
restore and the config migration refuse the write and leave the file as it
was. Fixtures encode through encodingtest.MustEncode.
Readers that hold only part of a stream say so rather than inherit the
whole-file answer:
- DetectFragment runs the round trip on the prefix the decoder consumes when
not at EOF, so a character the cut split is dropped structurally. grep's
8 KiB peek and read_file's 256 KiB sample use it.
- DecodeOutput is the one entry point for bytes that are only displayed.
The producer says which ends its bound cut (Cut{Head, Tail}); UTF-8 is read
without a character split at a cut end, and an uncut end keeps its bytes,
since a GBK pair there is output rather than half a rune. What no charset
restores is still read as GB18030. Hook output (tail cut when truncated),
the shell tail (head cut once it drops bytes) and the shell's combined
output (uncut) call only it.
- DetectAndDecode decodes a legacy charset once; the edit tools and
DecodeToUTF8 use it instead of detecting and decoding separately.
Tests that fail before this change (on the base or an earlier revision of it)
and pass here:
- TestE2EEditKeepsBytesOfUTF8FileWithStrayByte: e4 b8 ad became
e4 b8 84 31 a4 37 outside the edit (fails on origin/studio).
- TestDetectUTF8WithStrayByteIsNotGB18030 (fails on origin/studio).
- TestWritesRefuseCharacterCP936CannotHold (edit_file, multi_edit,
write_file): nil error and the whole CP936 file rewritten as UTF-8.
- TestDecodeOutputTrimsCutUTF8Edges: UTF-8 cut at the front read as mojibake.
- TestDecodeOutputKeepsUncutCodePageEdges, TestDecodeHookOutputKeepsUncutCodePageEdges:
GBK pairs at an uncut edge (bc db, b0 a1 0d 0a, 6f 6b 20 e4 a1,
b0 a1 20 6f 6b) were dropped as partial UTF-8 runes.
- TestEditKeepsCP936FileWithEuroByte: an edit adding Chinese wrote UTF-8
into a CP936 file.
- TestGrepGB18030PeekWithoutNewline, TestGrepGB18030PeekCutMidCharacter:
no matches past a peek cut mid-character.
- TestReadFileGB18030SampleWithoutNewline: garbled text past the sample.
- TestDecodeHookOutputReadsTruncatedCodePageText,
TestDecodeHookOutputReadsCP936Euro, TestCodePageOutputCutMidCharacterIsReadable:
mojibake for cut or euro-carrying cmd.exe output.
- TestEncodeRefusesCharacterTheCharsetCannotHold, TestDetectGBKWithEuroByte,
TestDetectFragmentDropsTheSplitCharacter, TestDecodeOutputReadsCutCodePageText:
the API's contract.
A balance_url pointing at https://openrouter.ai/api/v1/credits left the
balance readout empty. The wallet decoder is chosen by the endpoint's
host, and openrouter.ai had no entry, so the body was read as DeepSeek's
shape; {"data":{"total_credits","total_usage"}} carries no balance_infos
and was rejected as unreadable.
Add an openrouter.ai decoder: remaining = total_credits - total_usage,
in USD, available while positive. A body missing either field stays
ErrUnreadable rather than rendering as a zero balance.
Test: TestOpenRouterCreditsDecode and TestOpenRouterSpentWalletIsUnavailable
fail on the base ("wallet response not in this endpoint's shape") and pass
with the decoder; go test ./internal/model/billing/ is green.
`reasonix run` rejected --yolo and --dangerously-skip-permissions with
"unknown flag", `reasonix -y run "task"` failed with "unknown command",
and `reasonix --permission-mode auto run "task"` opened the terminal UI
with "run task" as its prompt instead of running headless. Print mode
disagreed with itself: `--yolo run -p hi` ran the task "run hi" and
`-y -p hi` was an unknown command.
Root cause: run's flag set never registered the yolo spellings the
terminal UI takes, and normalizeCommand only looked at argv[0], so a
leading session flag sent everything to the terminal UI, -y was read as
the command name, and the leading -p scan ran past the verb.
Fix:
- run registers --yolo/--dangerously-skip-permissions as an alias for
--permission-mode bypassPermissions; each conflict (-y with --yolo,
either with --permission-mode) names only the flags involved.
- run's flags move into runFlags/newRunFlags so they can be consulted
before dispatch.
- The verb is located with the union of the terminal UI's flags and
run's on the resume-normalized argv, as runTUI reads it, so a value of
either is never taken for it. Leading flags move after `run` only when
the session flags (the terminal UI's plus -y and -p) and run's both
parse them to the same flags with the same values. Otherwise the line stays the terminal
UI's: `--resume run hi`, `-r run hi` and `--inline run hi` keep their
meaning, and a run-only leading flag (--metrics, --output-format) is
reported by the terminal UI.
- -y and -p may lead like the other session flags, and the leading -p
scan stops at the verb even when a run-only flag precedes it.
- The first word after the leading flags is the verb, as it is with no
flags; CLI.md documents that, and interspersed run flags.
Test:
- TestRunApprovalFlagsReachTheHeadlessGate runs Run() against a fake
provider whose first reply calls write_file. The file appears only
where the resolved mode lets an unattended write through (ask and
dontAsk stay unwritten). On origin/studio the five new spellings exit
2 with no file. A mutation that drops the leading flags while routing
fails it for the three leading-flag spellings.
- TestNormalizeCommandCarriesLeadingFlagsIntoRun fails on origin/studio
(10 cases) and on the previous revision of this change (9 cases,
including the three terminal-UI regressions). The revision before this
one fails 2 cases: `--model m --metrics x run -p hi` and the same with
--output-format ran the task "run hi". The revision before this one
fails 2 more: `--resume --permission-mode
--permission-mode=bypassPermissions run hi` and `--resume --resume=abc
run hi` were forwarded to run although the two sets read different
values. It passes now.
- TestRunTakesTheYoloSpellings covers the alias and the per-flag
conflict messages.
Gates: gofmt, go vet (also GOOS=windows), make check clean,
golangci-lint 0 issues (also GOOS=windows), go test
./internal/frontend/cli/.
Refs #8355
Root cause: the todo_write schema declared "level" as
{"type":"integer","enum":[0,1]}. Gemini's function-declaration validator
accepts enum only on string properties with string members, so any Gemini
model behind an OpenAI-compatible relay rejected the whole request with
400 INVALID_ARGUMENT (TYPE_STRING on enum[0]/enum[1]), whatever the prompt.
Fix: express the same range as "minimum":0,"maximum":1. Execute already
rejects any level outside 0..1, so accepted input is unchanged. The level
description loses its "Nesting level:" lead-in to stay inside the schema
token budget. Boot golden fixtures are regenerated for the new bytes.
Test: TestBuiltinSchemaEnumsAreStrings walks every registered built-in
schema and fails on an enum that is not on a string property or has a
non-string member. On origin/studio it fails with three findings on
todo_write's level; with the fix it passes, as do
go test ./internal/tools/builtin/ ./internal/assembly/boot/.
Fixes#9005
A Responses turn could end with a tool call the model issued and never
run it. The stream reader dispatched a function call only when its
function_call_arguments.done or output_item.done event arrived; a relay
that forwards the added and delta events but not the done events (or
sends deltas without the item id) left the call open, the terminal
response.completed ended the attempt, and the turn committed as a clean
answer with no calls. The run then finished ok after one round.
The terminal response's output array is the response's own list of what
it issued. On response.completed and response.incomplete, every
function_call item there that the stream did not already close (matched
by call id) is now completed from that item; one whose status is set and
is not "completed" is left alone, so a call truncated at the output
limit is still not dispatched, and one listed without a call id is not
dispatched either.
The output array decodes one item at a time. Decoded as a typed slice, a
single item with a non-string field (arguments sent as an object by a
non-conforming relay) would fail the whole terminal event, which is then
skipped and the turn ends with unexpected EOF where it used to finish.
A call with no name, such as an arguments.done event that named no
item, is no longer dispatched as an empty call; it stays open for a
later event to name.
Test: wire_gap_test.go drives the real HTTP stream reader. On
origin/studio the "done events never came" and "deltas without item id"
cases fail with calls = nil; both pass after the change, and the
no-duplicate and truncated-call guards pass on both. The
unreadable-item, empty-call-id and unnamed-arguments.done cases fail
before the per-item decode and the skips, and pass after.
go test ./internal/model/... ./internal/runtime/agent/ passes.
The full-screen transcript showed an answer's thinking only as a
"thought for Ns" marker, with no way to read it; a long shell output
could already be opened with a click, the thinking could not.
Full screen, the marker now reads ▸ while shut and ▾ while open, with a
(Ctrl+O) hint. A left click on it, or Ctrl+O for the newest one (for a
terminal whose mouse is handed back to it), opens and shuts the thinking
underneath. Each opened row carries its own faint style, since the
transcript is split into rows after rendering. A streamed answer's first
chunk keeps its row, and when the answer settles that row takes the
final thinking and elapsed time, so thinking that kept arriving after
the text started is not lost. Inline mode, where printed rows cannot be
redrawn, keeps the fixed ▎ marker.
Tests (internal/frontend/tui/screen_test.go):
- TestThinkingOpensAndShutsOnAClick: streamed and settled answers open
and shut on a click of the marker row (FAIL on origin/studio).
- TestOpenedThinkingIsFaintOnEveryRow: every opened row starts with the
faint SGR under an ANSI256 profile (FAIL with one style around the block).
- TestStreamedThinkingOpensAsItSettled: thinking arriving after the text
and the final "thought for 4s" show when opened (FAIL with the row
copied at the first chunk).
- TestCtrlOOpensTheLatestThinking: Ctrl+O opens and shuts it with the
mouse handed to the terminal (FAIL without the key).
Fixes#7934
Root cause: POST /providers/edit saved the entry and answered 204 without
rebuilding the runtime. The context window, token limits, headers and
reasoning protocol are bound into the agent at assembly, so the open
conversation kept its old window: the gauge and the fold point went on
counting against the number the form had just replaced, until something
unrelated (a model switch) rebuilt it. /context/window already rebuilds
in place; the edit form did not.
Fix: after the save, act only when the conversation runs on the edited
source and a field the agent binds at assembly actually changed (base URL,
models, vision, per-model overrides, window, max output, headers, extra
body, protocol, efforts). A key-only edit needs no rebuild because the key
is read per request, so it answers 204. Otherwise:
- running model still listed: rebuild in place. With work in progress (a
running turn, a pending question or background jobs) the save still
lands and the answer is provider.saved_while_running; any other rebuild
failure is runtime.rebuild_failed.
- the edit unticks the running model: a plain save, 204.
- the running model was already unlisted before the edit: nothing on the
entry can carry the change to that conversation, so the answer is
provider.saved_model_unlisted ("switch models to apply it").
On any of these "saved but not applied" answers the edit form refreshes
the providers list and stays open to show why.
Test:
- TestEditProviderReachesTheConversationRunningOnIt failed on
origin/studio ("window = 0, want the saved 1000000"); passes.
- TestEditProviderMidTurnSaysTheConversationKeepsItsSettings failed on
origin/studio ("mid-turn = 204, want 409"); passes.
- TestEditProviderDroppingTheRunningModelIsStillSaved and
TestEditProviderKeyOnlyMidTurnNeedsNoRebuild failed on an earlier
revision of this change (409 instead of 204); pass.
- TestEditProviderAfterDroppingTheRunningModelSaysItIsUnlisted failed on
the previous head ("editing the window after the drop = 204, want
409"); passes.
- EditConn test: a saved_while_running answer calls onSaved, not onDone.
The serve package, go vet (also GOOS=windows), golangci-lint, make check,
tsc, vitest, census:verify and perf/i18n.mjs pass.
A task, parallel_tasks or fleet call may name a child model without a
provider, e.g. "deepseek-v4-flash". The sub-agent resolver passed that
straight to Config.ResolveModel, which answers a bare name with the first
provider that lists it. With the official DeepSeek entry listed before a
relay that serves the same model, a child of a session running on the relay
was silently sent to the official API instead.
A bare name the parent's provider serves is now qualified with that
provider before resolution, in both the runtime resolver and the identity
the usage record reports. A provider name, a provider/model reference, or
a model only another provider lists resolve as before.
Test: TestSubagentBareModelStaysOnParentProvider fails on origin/studio
(resolved refs = ["first/shared-flash" ...], want "relay/shared-flash")
and passes after the fix; go test ./internal/assembly/boot/
./internal/runtime/delegation/ pass, make check clean.
A reopened Studio conversation showed the host's own instructions to the
model as if the user had typed them: the tool-round / task-budget
finalization request, the retry after a reasoning-only answer, the
executor-handoff retry, and the output-truncation fact.
Root cause: /history and the TUI restore decide "who said this" from
provider.Message.HostAuthored, which is written by the producer. Four
injection sites in the run loop appended role=user messages without it
(only the workflow-phase repair set it), so the transcript rebuilt them
as user bubbles and session previews could count them as user turns.
Fix: set HostAuthored at each site. Provider requests are unchanged:
the projection clears the field before sending.
Test: TestLoopComposedUserMessagesAreHostAuthored drives a real Run into
the task-budget landing, an empty-final retry and a truncated tool call.
Before the fix all three cases fail ("a line the loop composed is
recorded as the user's"); after it they pass. agent, serve, control,
sessionstore and boot packages pass; make check and golangci-lint clean.
Fixes#8486
A code block's copy button already swaps to a check mark after a copy,
but hovering it said "Copy response", and the config-trouble banner's
button for the config file path read "Copy response" as its visible text.
Root cause: CopyButton's idle label was the fixed string for a chat
reply; the caller's `label` only reached aria-label, so title and
visible text ignored it. ConfigTrouble passed no label at all.
Fix: the idle label is the caller's `label` when given, and the config
banner labels its button "Copy path" (en locale entry added).
Test: src/ui/CopyButton.test.tsx fails on origin/studio (code block
title "复制回复" vs "复制这段代码"; path button text "复制回复" vs
"复制路径") and passes after. Full vitest (928), tsc, census:verify
and make check pass.
Choosing the free-text option on an ask card left the caret on the
option button, so the answer could not be typed until the box was
clicked, and Enter in the box did nothing: every typed answer needed
the mouse twice.
Choosing the option now focuses its box, every time it is chosen, and
Enter there does what the primary button would. When Enter moves to the
next question, focus follows into that pane (its open box, else its
first option), because the pane it leaves becomes display:none. The key
goes through useIme, so the Enter that confirms an IME candidate is not
read as an answer. The box carries an aria-label.
Test (PromptCards.interaction):
- the caret lands in the box after choosing the option; a composing
Enter does not answer, a later Enter does (clock stubbed, not slept);
- Other -> a normal option -> Other again puts the caret back in the
box (FAIL before: focus stayed on the button);
- Enter on question 1 moves focus to question 2's first option (FAIL
before: focus stayed in the hidden pane's input).
Before/after screenshots were taken on the perf build, light and dark.
Fixes#7904
An ACP client that sent a picture got an answer from a model that never
saw it. session/prompt flattened its blocks to text and kept only text
blocks and resources with inline text: an embedded resource carrying a
base64 blob, which a client sends because embeddedContext is advertised,
was dropped without a word, and so was any image block. An image-only
prompt was refused as empty.
Image blocks and blob resources are now decoded and stored with the
control package's attachment savers under .reasonix/attachments/ in the
session's workspace, and the prompt carries the "@path" reference that
serve and the CLI already hand a turn. Undecodable data or bytes that do
not sniff as an image fail the prompt with invalid params instead of
vanishing. Steer uses the same path. initialize now advertises
promptCapabilities.image = true.
Test: prompt_attachments_test.go drives session/prompt through the real
Serve loop and a real controller. On origin/studio the image block and
blob resource cases fail with the turn input "what is in this picture?"
(no reference), the image-only case fails with "empty prompt", and the
undecodable case returns no error; all pass after the change, and
go test ./internal/frontend/acp/ passes.
On a Windows host without Git Bash the shell tool resolves to PowerShell,
a supported path. Five tests still assumed bash there:
- TestBackgroundBashWaitAndOutput, TestBashTerminalFallsBackWhenUnhandled,
TestBashTerminalSkippedWhenEnvFilteringEnabled and TestRunForegroundSuccess
ran `printf`, which PowerShell does not have. They test job, terminal and
runner plumbing, not POSIX syntax, so they now run `echo`, which both
interpreters accept with the same output.
- TestDefaultSpawnerUsesGitBashForExplicitShOnWindows needs a real Git Bash;
it now skips when no bash is installed. Discovery itself stays covered by
the injected-host tests in internal/safety/sandbox.
- TestGoldenBaselineNoExtensions records the bash tool's bash description.
A host that resolves PowerShell renders a different, intended description,
so the golden now skips there instead of reporting drift, and is never
regenerated from such a host.
Test: the touched tests pass on macOS (bash host) and the four packages pass
in full; go vet passes with GOOS=windows. The PowerShell-only failure could
not be reproduced here (no Windows host without Git Bash is available); the
reporter's reproduction in #8326 shows `printf` not recognized and the
golden diff on the bash tool description.
Problem: a long high-effort think on Studio ended with "response stopped:
hit the client reasoning safety limit" and no answer, even though the
provider was still inside the output budget it had been sent.
Root cause: the agent cancelled the provider stream once 128 KiB of hidden
reasoning had been buffered (about 32K estimated tokens) and returned
errReasoningByteLimitExceeded. The request already carries a provider
output budget (auto ~64K on DeepSeek high), so the byte guard fired inside
the budget and turned a legitimate think into a failed turn. The 1.x line
dropped the abort in #8779; Studio never received it.
Fix: the byte limit now bounds only what the host retains. Reasoning keeps
streaming live, the stored copy is the rune-safe prefix of the stream's
first ReasoningByteLimit bytes, and the default rises to 8 MiB so no
provider-bound reasoning is cut under current output ceilings. Interrupted
usage estimates count every streamed reasoning byte. The sentinel and the
client_reasoning_limit finish reason had no producer left and are removed.
Test: TestDefaultReasoningLimitLetsLongThinkingFinish and
TestReasoningByteLimitBoundsRetainedReasoningOnly fail on origin/studio
with "reasoning output exceeded client byte limit" and pass after the fix.
go test ./internal/runtime/agent ./internal/assembly/boot
./internal/tools/builtin, the race run of the reasoning tests, go vet,
golangci-lint and make check are clean.
Hooks could see a sub-agent finish (SubagentStop) but not begin, and
SubagentStop fired only when the child answered. A failed, cancelled or
self-refused task ended silently, so an external observer driven by
hooks (activity indicators, notifiers) had no event marking the start
of delegated work and no reliable end.
SubagentStart is a new non-blocking event fired just before a
foreground `task` call executes, with the call's arguments in toolArgs.
SubagentStop now fires from a defer on every exit after that start
(answer, failure, cancel, refusal) with `error` set when there was no
answer, under a context detached from the turn's cancellation so a
cancelled turn does not kill the hook. Both carry the call id as callId
(tool_use_id in the Claude-format payload) so consumers can pair them.
Only a foreground `task` fires them; read_only_task, parallel_tasks,
fleet, skill sub-agents and background tasks fire neither. hook.go,
the reasonix-guide skill and GUIDE (en/zh) state that scope and that
exit 2 cannot block.
Test: TestSubagentStartPairsWithSubagentStop runs answer, failure,
cancel and refusal cases plus a background call. Before (SubagentStop
on success only): fails, cancelled and refuses each fail with
"SubagentStart must be followed by exactly one SubagentStop, saw 0".
After: all pass, with matching call ids, err set on the three non-answer
cases, and a live context in SubagentStop. TestRunnerSubagentStartPayload
checks callId, toolArgs and error reach the hook's stdin.
A global [[plugins]] entry had no way to name the workspace it runs for:
${VAR} expansion read only the process environment and the project
.env. Servers that key on the project path in a header or argument
(IntelliJ's IJ_MCP_SERVER_PROJECT_PATH) therefore needed one entry per
project. Installed plugin packages already got ${CLAUDE_PROJECT_DIR};
user and .mcp.json entries did not.
ExpandedPluginForRoot resolves ${REASONIX_WORKSPACE_ROOT} and
${CLAUDE_PROJECT_DIR} to the workspace, and both places that turn an
entry into a spec call it with the root they already hold:
pluginspec.FromEntry (session boot, capdiag, mcp oauth) and the
controller's mcpIdentitySpec (live add, install, reconnect). The host
owns these values, so they outrank a process or .env variable of the
same name; a hook's child process inherits its parent session's
CLAUDE_PROJECT_DIR. Config path fields (sandbox roots, skill paths,
proxy) keep plain environment expansion. WorkspaceRootValue is the one
canonical form (absolute, cleaned, symlinks as opened), also used for
the plugin-package substitution.
Tests (the first three fail before the fix and pass after):
- TestAddMCPServerExpandsWorkspaceRootInLiveHeaders: live AddMCPServer
against an httptest MCP server. Before: live headers = ("",
"/inherited/from/a/parent/hook").
- TestLoadForRootExpandsWorkspaceRootInPluginEntries: two workspaces,
one global http entry and one .mcp.json entry, with a conflicting
CLAUDE_PROJECT_DIR in process env and .env. Before: header = "".
- TestPluginSpecsForRootExpandWorkspaceRoot covers the boot mapping, and
TestWorkspaceRootVarsStayOutOfConfigPathExpansion keeps sandbox paths
out of scope.
Typing "@proinf" found nothing for ProjectInfoNew.tsx: fileref.Search
only recorded a hit when the query was a contiguous substring of a
directory name, the basename or a path segment, so a file could only be
reached by spelling a run of its name exactly or walking down to it one
directory at a time.
Search now keeps a last tier for names that contain the query's letters
in order (case-insensitive), files by basename and directories by name.
It ranks after every contiguous hit and only fills the slots those
leave, so existing results and their order are unchanged. The kernel's
Complete and the external-folder ref search both go through Search, so
every frontend gets it.
Tests: TestSearchMatchesSubsequenceAfterSubstringHits (fileref) and
TestCompleteRefFindsFileBySubsequence (control.Complete, the boundary the
composer calls) fail on origin/studio (got only [src/proinfo.go]; items
= []) and pass with the change.
Under a fake-ip resolver (Clash / mihomo) every public host answers with a
routable A from 198.18.0.0/15 beside a unique-local AAAA from fd00::/8.
The direct dial path refused the whole host as soon as any resolved address
was on the SSRF blocklist, so web_fetch failed for every site with
"refusing to fetch internal address ... (resolves to fdfe:...)".
Root cause: the guard treated one blocked address as a verdict on the host,
although the dial is pinned to a resolved address rather than the name. A
blocked address in a mixed answer can simply be skipped; the connect still
only ever reaches an address that passed the check. A host whose addresses
are all blocked is refused exactly as before.
The direct path moves into dialVettedAddress, which tries the allowed
addresses in resolver order, and the lookup becomes a package seam so the
test can supply a resolver answer.
Test: TestWebFetchDialsAllowedAddressBesideBlockedOne resolves a name to
[fdfe:dcba:9876::152, 127.0.0.1] and fetches from a loopback server. On the
base it fails with "refusing to fetch internal address api-docs.example
(resolves to fdfe:dcba:9876::152)"; with the fix it passes.
TestWebFetchRefusesHostWithOnlyBlockedAddresses keeps the all-blocked
refusal. go test ./internal/tools/builtin/, go vet, make check and
golangci-lint are clean.
A local model running at ~25 tok/s can spend well over two minutes
prefilling a long context without sending a byte. The openai, anthropic
and responses providers each cut such a call after 120s of silence (the
SSE idle watchdog, and the transport's response-header timeout on the two
that set one; anthropic set none, so its header wait was unbounded), and
the turn failed with "stream stalled — no data for 2m0s" although the
model was still working. The 1.x line already moved these guards to 300s.
The window now lives once, as provider.StreamIdleTimeout (300s), and all
three providers read it for both the idle watchdog and the header wait.
A longer header window multiplies through SendWithRetry, which retries a
response-header timeout with the full MaxRetries budget: 11 x 300s plus
backoff is close to an hour before the user sees an error, and a
prefill that timed out once times out identically on every attempt. A
transport timeout (a net.Error reporting Timeout() while the caller's
context is still live) now gets its own budget of one retry, the way
maxAuthRetries bounds auth failures, so the worst case is about ten
minutes.
Test:
- TestNewWaitsFiveMinutesForASilentModel in openai, anthropic and
responses builds each client through New and reads its idle and
header windows. Before: "stream idle window = 2m0s, want at least 5m"
in all three, and "response header window = 0s" for anthropic.
- TestSendWithRetryGivesAHeaderTimeoutOneRetry drives a real
http.Transport against a server that never sends headers. Before:
"calls = 4, want the timed-out attempt plus exactly one retry" (retry
limit 3). After: two calls and an error that is a net.Error timeout.
Fixes#6990
A Windows machine whose OpenSSH DefaultShell is Git Bash could not be used
as a remote host: `uname -sm` answers "MSYS_NT-10.0-26200 x86_64", which is
neither Linux nor macOS, and the fallback `echo %OS% %PROCESSOR_ARCHITECTURE%`
is cmd syntax that bash (and PowerShell) print back literally, so the machine
was refused as unsupported.
Root cause: detectPlatform asked the Windows question in cmd's syntax, while
every command windowsShell sends afterwards already goes through
`powershell -EncodedCommand`, which any of the three login shells runs the
same way.
Fix: ask for OS and PROCESSOR_ARCHITECTURE through psCommand. The guide no
longer says remote hosts are Linux and macOS only.
Test: TestWindowsIsIdentifiedWhateverTheLoginShell drives detectPlatform
through cmd, PowerShell, Git Bash (MSYS_NT) and MINGW64 login shells. On
origin/studio the Git Bash case fails with
bootstrap: unsupported remote: `echo %OS%` answered "%OS% %PROCESSOR_ARCHITECTURE%"
and all four pass after the change.
The explorer classified entries by os.ReadDir's DirEntry.IsDir, which is
false for any symlink. A link to a folder inside the workspace was drawn
as a file, so it could not be expanded, and clicking it asked for a file
read that fails. Search had the same flaw: WalkDir does not follow links,
so a linked folder whose name matched came back as a file hit, and a link
to a file outside the tree came back as a hit the file endpoint's
os.OpenRoot then refuses.
A link entry is now resolved: one naming a folder inside the tree is
listed as a folder (expanding it already works, listFolder checks the
real path), one naming a file inside the tree as a file, and one that
dangles or leaves the tree is not offered, matching the refusal
listFolder already gives such a folder.
Test: TestALinkIsListedAsWhatItResolvesTo and
TestASearchListsOnlyLinksToFilesInTheTree fail on origin/studio
("directories = [alpha zeta]", "search = [alpha/a.txt far.txt near.txt
txt-folder zeta/z.txt]") and pass with the fix; the serve package
passes in full.
Refs #5914
The planner keeps a session of its own and only ever saw its own plans,
so the next planning turn did not know what the executor had found or
changed, and turns the policy routed straight to the executor never
reached it at all. Users ended up relaying the executor's results to the
planner by hand.
Every executor run now owes the planner its final reply (plus the user's
input when the planner never saw that turn), and the next planning input
carries the owed outcomes in an <executor-since-last-plan> block at the
turn tail, so the planner's cache-stable prefix is untouched. The debt
clears only once a plan comes back; a planning turn that fails and rolls
back keeps it owed. Switching sessions clears it with the planner
transcript.
The debt is bounded where it is held: task and reply are cut to their
render caps on append, only the newest four outcomes are kept, and a
count of the trimmed ones is rendered. A reply is taken only from the
run's own messages; when a compaction rewrites the session mid-run,
those begin at the run's own user message, identified by a stamp later
than every message before the run, and without it the outcome reads
"no reply captured" instead of an earlier turn's reply.
Test: TestPlannerSeesWhatTheExecutorDidSinceItLastPlanned and
TestPlannerKeepsExecutorOutcomeOwedWhenPlanningFails fail on
origin/studio (the second planning input carries no executor reply).
TestExecutorOutcomeNeverReportsAnEarlierReplyAfterCompaction (an
earlier reply reported twice) and TestOwedExecutorOutcomesStayBounded
(7 outcomes held, want 4) fail on the previous revision. All pass after,
with the existing coordinator suite.
complete_step rejected diff/files evidence whose paths were cited
workspace-relative ("frontend/src/core/backend/idb.ts") when the writer
had been given the absolute path: the ledger records a receipt's path as
the tool received it, and complete_step compared cited and recorded
paths for exact equality, in the per-turn ledger and in the cross-turn
session fallback alike. The step could not be signed off even though
the file had been written.
The agent now attaches its workspace root to each tool call's evidence
scope, and complete_step checks every cited path under the forms that
name the same workspace file: the citation itself, a relative citation
joined to the root, and an absolute citation made relative to it. A path
outside the workspace keeps only its own form, relative-vs-relative stays
exact, and case folds only on Windows through NormalizePath. The ledger
queries themselves are unchanged, so completion reports and blocked
conclusions keep exact matching.
Test: TestCompleteStepCreditsCitationsOfTheSameWorkspaceFile and
TestE2ERelativeCitationOfAbsoluteWriteSignsOff fail on origin/studio
("diff paths have no matching successful writer receipt") and pass
after. TestCompleteStepCreditsNoOtherFileForACitation (a same tail
outside the workspace, a base name, a relative nested write, a deeper
file's trailing components, a .bak, a ../ escape, a different case on
Linux) is rejected on origin/studio, credited under suffix matching,
and rejected after.
A workspace below a Box Drive mount on Windows failed every config load
with "resolve project config path ...: The system cannot find the path
specified", whether or not reasonix.toml existed.
Root cause: resolveConfigAccessPath canonicalizes the whole path with
filepath.EvalSymlinks before it confines a project config link to its
directory. Below a cloud-drive reparse point that canonicalization fails
while the files themselves open normally, and the failure was treated as
fatal even when reasonix.toml was a plain file or absent.
Fix: when canonicalization fails and the config entry itself is absent or
a non-link file, use the logical path. Such a file lies in its own
directory whatever its ancestors resolve to, so the confinement check has
nothing to decide. A link still has to resolve, so an unresolvable link is
refused as before.
Test: path_access_test.go stubs canonicalization to fail below a mount
directory. Without the fallback both LoadForRoot cases fail with the
reported "resolve project config path" error; with it they load (and the
project value is applied), and a reasonix.toml link under the same mount
is still refused.
Off macOS the Electron shell sets the application menu to null so no bar
renders inside the window. Electron binds F11 only through a menu role
(togglefullscreen), so with no menu the key did nothing and the window
had no way into full screen on Linux or Windows.
The shell now listens for a bare, non-repeated F11 keyDown on the main
window's contents and toggles setFullScreen. macOS is left alone: it
keeps full screen on the title bar and the window menu.
Test: desktop/electron/test/unit.mjs "F11 toggles full screen where no
application menu binds it" fails on origin/studio (the key is unbound)
and passes with this change; node --test test/unit.mjs is 43/43.
The environment block in the system prompt printed "- Shell: bash". That
value is the interpreter sandbox.ResolveShell picked for the bash tool; the
user's login shell ($SHELL) never feeds it (it is only read to recover
PATH). Unattributed, the line reads as the user's shell, so a fish or zsh
user got advice about ~/.bashrc.
The line now reads "- Shell used by the bash tool: <shell>", and the
rendered [tools.shell] config section says the setting picks the bash
tool's interpreter and is independent of the login shell. The prefix
changes once on upgrade and stays byte-stable within a session.
Test: TestBuildEnvironmentBlockAttributesShellToBashTool builds through
boot with SHELL=/usr/bin/fish and reads the system message sent to the
provider. On origin/studio it fails (block shows "- Shell: bash"); with
the fix it passes, together with the environment and config packages.
Fixes#6823
The config documents ui.show_turn_usage as the switch for the per-request
"TURN … tok · in … · cached …" receipt in the CLI/TUI transcript, and the
1.x TUI reads it. The studio TUI never did: runTUI loaded the config but
passed nothing about it to tui.Options, and every usage row was drawn, so
setting it to false changed nothing.
tui.Options gains HideTurnUsage, set from the loaded config. A hidden usage
row still folds into the transcript (a later frame of the same attempt
restates it in place and TurnOut keeps counting), but it is marked committed
without being printed and is skipped in the live area.
Test: TestHiddenTurnUsageNeverReachesTheScreen drives a turn through the
model and checks the screen while the answer streams and after it settles.
Before the fix, hide=true failed in both states (receipt shown); after it,
both pass, and hide=false still shows the receipt.
Root cause: the workbench read an open file once, keyed on its path only.
The explorer tree beside it is re-read when a turn settles, when the
window is looked at again and when the change set moves, but the file
view never was, so an edit by the agent or another application stayed
invisible until the tab was closed and reopened.
Fix: the file view is re-read on the same occasions as the tree, and the
content is replaced only when the kernel's revision for it moved. A draft
with unsaved edits is never overwritten, including one made while a
reload read is in flight, and a hidden panel reads nothing. Revisions
are opaque, so a read only counts if no save and no later read has been
issued since it started; otherwise a reload issued before a save could
land after it and revert the view to the pre-save content.
Test: WorkbenchPanel.test.tsx adds five cases.
- "shows an open file as it is on disk once a turn settles" and
"... once the window is looked at again" fail on the base (view still
shows "First") and pass after the fix.
- "keeps unsaved edits when the file changes on disk underneath them"
pins that a dirty draft issues no read.
- "keeps an edit made while a reload was reading the file" fails when the
in-flight dirty-draft guard is removed and passes with it.
- "drops a reload that was issued before a save and lands after it"
fails without the read/save ordering (view reverts to "# Second") and
passes with it.
Full frontend-next vitest (915 tests), tsc --noEmit and make check are
clean.
Refs #5693
A fenced code block is drawn with a "│ " rail in its gutter. With the
mouse captured, selectedText read the rows exactly as drawn, so copying
code from an answer pasted a rail at the start of every line. Native
mouse mode already draws the rail as blank cells; the in-app copy now
reads the transcript rendered that way. The rail and its blank form have
the same width, so rows and columns still line up with the selection
drawn on screen, and only rows through the selection's end are rendered.
Test: TestSelectionCopiesCodeWithoutTheRail selects a railed code block
in a settled and a live answer. Before: copied
" │ func a() {\n │ return\n │ }". After: no rail; the tui package,
go vet, golangci-lint and make check pass.
remark-gfm defaults to singleTilde: true, so "500~1000 or 2000~3000"
struck out everything from the first tilde to the second. A lone tilde is
the ordinary way to write a range, especially in Chinese replies. The
transcript renderer now passes singleTilde: false; ~~text~~ still strikes.
Test: Markdown.plugins.test.tsx renders two ranges and one ~~...~~ span.
Before the fix it found 2 <del> elements; after, exactly one, holding
only the doubled-tilde text. Full frontend-next vitest and tsc pass.
SendWithRetry retries transient transport errors, but transientErr refused
anything matching errors.Is(err, context.DeadlineExceeded). net/http's
"timeout awaiting response headers" error implements Is(DeadlineExceeded),
so a slow gateway that hit ResponseHeaderTimeout ended the request on the
first attempt and left an unattended run stopped.
Whether the caller gave up is now asked of the caller's context rather than
read off the error's identity; a transport timeout while the caller is still
waiting is retried with the usual backoff.
Test: TestSendWithRetryRetriesResponseHeaderTimeout drives a real
http.Transport with ResponseHeaderTimeout against an httptest server that
stalls the first request. Before: "p: request failed: ... net/http: timeout
awaiting response headers". After: recovers on the second attempt.
TestSendWithRetryStopsWhenCallerDeadlineExpires pins that the caller's own
deadline is still not retried.
Refs #6354
A model switch rebuilds the controller from a snapshot of its history and
closes the old one. /compact runs asynchronously under the rotation gate,
but RuntimeStatus reported only running/finishing as active, so every
busy guard built on it (serve switchModel, provider/workspace rebinds,
ACP config rebuilds) let the switch through mid-fold: the new controller
carried the pre-fold history while the fold landed on the discarded
controller, which then wrote the compacted transcript over the same file.
RuntimeStatus.Running now includes the rotation gate, which also covers
rewind, summarize and new/clear session, matching the 1.x line where
maintenance already counts as running.
Test: TestSwitchModelRefusesWhileManualCompactIsFolding holds a real
manual compaction open at its summary call and switches the model through
serve. Before: switchModel returned nil and replaced the controller.
After: it refuses with busy.switch_model and keeps the old controller.
The Windows smoke and full legs looked hung: output stopped before
internal/assembly/boot and the step ran into its job timeout without
-timeout ever firing. Nothing was hung. With the result cache on, go test
passes each test binary -test.testlogfile, and before and after the run
it computes the cache key from that log, calling filepath.EvalSymlinks on
every recorded path outside the module. boot's log holds about 1.9M
entries (84 MB); on a Windows runner that costs about five minutes per
pass, and a restored cache that already holds a boot log pays it twice.
Pass -count=1 to every root-module go test that runs on Windows, and hold
ci.yml to it with a test that reads the workflow.
A quoted paragraph was wrapped to the full terminal width and then
shifted right by the "▎ " rail (plus any list indent), so every full
row ran two or more columns past the edge. The terminal soft-wrapped
the overflow, leaving a stray rail on one row and the last characters
of the line on the next — most visible with CJK text, where every row
is filled to the edge.
renderBlockquote now lays the quote's body out at the width left after
its prefix and restores the renderer's width afterwards, so nested
quotes and quotes inside list items narrow once per level.
Test: TestBlockquoteRowsFitTheRenderWidth renders top-level, nested,
list-indented and Latin blockquotes at widths 20/41/80 and asserts no
row exceeds the width. It fails on the base (an 84-column row at width 80, 24 at width 20)
and passes with the fix.
A reader who audits what the agent wrote had no way to keep diffs open
while leaving reads and commands folded. The folding settings answered per
part of the transcript (work group, step) and never per what a step did:
"steps" opened everything or only failures, and the work group folded once
the turn ended, so an edit was always two clicks deep.
Folding gains a "changed" mode for the work group and for steps. A step
opens when it changed a file (it carries a diff or line counts, the same
fact the +N/-M badge already reads); a work group opens when one of its
calls did. A "Changes only" preset sets both. The settings preview gains an
edit step so the mode shows what it does.
Test: src/ui/fold-changes.test.tsx renders a real Transcript with a read-only
turn stretch and one holding bash + edit_file. On the base, "changed" is
not a stored mode and both groups stay folded (expected [false, true],
got [false, false]); with the fix only the edit's group and the edit's
step open. Full vitest (144 files, 912 tests) and tsc pass.
curl, wget, pip and most download tools redraw their progress meter by
returning to column zero with a carriage return. The tool card renders
shell output in a <pre>, and a browser draws a carriage return as a
space, so every frame the command printed landed side by side on one
line. The card showed the first frame (0%) and the rest ran off to the
right: a 64-frame curl download measured 35003px wide in Chromium, so a
running download looked stuck at its start.
Term now resolves each line the way a terminal would leave it: every
frame overwrites the start of the one before, a CRLF ending stays a line
ending, and a trailing return holds the last complete frame while the
next one is still arriving. Copy still takes the raw output.
Test: four cases in ToolCards.state.test.tsx (latest frame shown,
shorter frame keeps the uncovered tail, CRLF, trailing return) fail on
origin/studio and pass with the fix. In the perf build the same curl
card goes from 35003px to 710px wide and shows 100% instead of 0%.
Root cause: the full-screen transcript only bound plain PgUp/PgDn, so the
Shift+PgUp/PgDn a terminal uses for its own scrollback fell through to the
composer and did nothing. Shift+Insert, which Linux terminals that pass the
key through expect to paste, was unbound too (the textarea binds only Ctrl+V).
Ctrl+C quitting (twice, when idle) and Ctrl+Insert copying a selection were
already in place.
Shift+Up/Down stay with the composer, which uses them to select lines.
Test: TestShiftPageKeysScrollTheTranscript and
TestShiftInsertPastesClipboardText fail on origin/studio ("shift+pgup left
the view on the tail", "shift+insert did nothing") and pass with the fix;
go test ./internal/frontend/tui/ passes.
Refs #4912
Anything a task produced had to be found by hand under the workspace
path, because the explorer had no way to open Explorer or Finder at a
file or folder. Nothing like it existed. The only reveal in the app
opens the theme folder.
The explorer header gains a button that shows the workspace root. Every
file and folder row gains a context menu:
- "Reveal in File Explorer" on Windows
- "Reveal in Finder" on macOS
- "Show in file manager" elsewhere
The kernel confines the path and the shell does the showing.
- GET /workspace/locate?path=<workspace-relative> answers the absolute
path of an entry in that runtime's workspace. "" and "." name the
root. It refuses "..", absolute paths and links that resolve outside
the tree with workspace.path_outside_tree, and a missing entry with
workspace.file_missing.
- The route is gated on the local-desktop host grant. AllowEditorOpen
is renamed AllowLocalDesktop because the grant now covers both
routes. A networked serve and a paired device get
workspace.locate_no_window.
- The hub refuses the route for a remote pane before proxying, with the
same code. Only the hub knows the pane is remote, and a far kernel,
compromised or not, answers with paths that mean nothing here.
- The renderer sends only its pane base and the relative path, over the
new shell:reveal IPC. Electron main does the rest:
- it requires the base to be a pane (/rt/<id>);
- it asks the kernel with its own credential;
- it acts only on an absolute path the kernel answered, and on Windows
only on a drive-letter path, never a share or device path
(\\host\share, \\?\, \\.\, \??\);
- it only ever calls shell.showItemInFolder, the root included.
openPath is never used: a directory can be an application bundle,
or a link to one, and opening it would launch it outside the
sandbox.
- StudioHost now forwards the query string, which it used to drop
silently.
- The page hides the action in a browser tab (no bridge) and in a
remote pane.
Tests:
- workspace_locate_test.go:
- entries inside the tree, and "." as the root;
- escapes, including a symlink escape;
- a missing entry;
- no grant, and a paired device;
- the hub refusing a remote pane without reaching the far kernel.
- Electron unit tests drive reveal() through StudioHost against a real
HTTP server. They check that the root is only selected, even when it
is answered as a directory, that forged bases never reach the kernel,
and that relative answers and Windows share or device answers are
refused.
- WorkbenchPanel tests cover the header button, both row menus, the
keyboard path, the remote pane and a coded refusal.
Show a pending state while slash commands and skills load, and keep a dismissed pending request from reopening the menu when it resolves. Covers the delayed and Esc paths with interaction tests. Fixes#9311.
In Settings > Model services the protocol row (.vway) lays out the label,
the segmented control and the hint in one nowrap flex row. The label and
the control are flex: none, and the hint is the only item that can shrink
(flex: 0 1 auto, min-width: 0). The detail column is only ~474px wide even
in a 1400px window: label 47 + control 367 + gaps 20 + padding 24 leave
16px, so the hint was shrunk to 16px and set one word per line, 24 lines
down the pane (computed styles read in Chromium). The only rule that
wrapped the row was a max-width: 760px viewport query, which never fires
while the pane itself is narrow inside a wide window.
A switcher row now always wraps, and the hint takes flex: 1 1 100%, so it
sits on its own line under the switcher at every width - the way the
neighbouring rows put a description under their label rather than beside
their control. The viewport query that did this only on narrow windows is
removed.
Test: perf/overflow.mjs opens Settings > Model services at 1400px and 640px
and measures every switcher hint in the detail pane (width against the
room in its row, rendered line count).
Before (origin/studio): 1400px FAIL - width 16px of 450px, 24 lines;
640px ok. After: 1400px 450/450px, 2 lines; 640px 538/538px, 2 lines.
tsc --noEmit, vitest src/ui/Providers src/ui/AddProvider
src/ui/settings-catalog src/styles (36 passed), perf/contrast.mjs clean,
make check clean.
A running bash step in the activity group showed the same static clock as
the group header; the only sign the turn was alive was the composer's R.
Each step row also left a wide blank between the status mark and the
command.
Root causes:
- Nothing animated a running call's state mark; running differed from
done only by a tint.
- `.activity-body .hl .nm` was `flex: 0 0 96px`: a fixed name column, so
a short name like `bash` left ~80px of empty box before the command.
- `.flow .call[data-k="say"] .hl` (and its `.nm`/`.arg` rules) matched
every descendant `.hl`. An activity group is re-hosted inside the say
card, so a running row (a plain `.hl`, not a disclosure summary) picked
up the answer header's 29px left padding and 10px bottom margin and sat
26px right of the finished rows; args also took `--muted` instead of
the group's `--faint`.
Fix:
- A running call's clock turns (2.4s linear) while its state holds. One
fact moves in one place: the step's own mark while the group is open,
the group header's mark once it is folded. The header also says
"{n} 项运行中" in `--net`. Reduced motion stops the turn and keeps the
tint; hidden panes pause it.
- The name hugs its text up to 96px and never shrinks.
- The say header rules are scoped to the card's own header
(`> .c > .hl`), and plain step heads get the summary's 3px inset.
The motion guard counted any two infinite loops as the whole budget; it
now classifies each loop by what it reports (the run mark, or a running
call's mark at most once per running call) and fails any other loop.
Test: new perf/running.mjs drives a finished and a running bash call via
__feed in dark/light at 1440px and 640px, plus reduced motion. Against
origin/studio it fails 28 checks (no animation on the running mark, no
header signal, 83px between name and command, rows 26px apart); after
the fix all pass. motion, idle, cards, contrast, overflow and surface
guards pass; vitest 908/908; tsc clean; make check clean.
The explorer kept listing folders that had been deleted in Windows
Explorer: 12 entries against the 4 the OS showed.
Root cause: GET /workspace/files reads the disk on every request and keeps
no cache, but the panel only asked when it was opened, when the search
query changed, or when the git change set changed. That change set is
itself fetched only at turn boundaries, and untracked or ignored folders
(and every folder of a workspace that is not a repository) never move it.
So a change made in another application was never followed by a read, and
the tree's freshness rested on the agent being the one who wrote.
Fix: the listing is also read again when the window regains focus or its
page becomes visible (useGlance), and when a turn settles. No fs watcher:
the kernel listing is uncached and cheap, a watcher would need a new
dependency, a handle per expanded folder on Windows, and a new pushed
event type across the wire, all to refresh a panel nobody is looking at.
Test: WorkbenchPanel.test.tsx "drops a folder deleted outside the app once
the window is looked at again" deletes a folder from the backing store
without any port call and fires window focus. Fails on origin/studio (the
t620 row is still present), passes with the fix. A second case covers the
turn-settle read of a file git does not list.
The clean delivery receipt lays out the tick, the label and the evidence
(file count plus passing commands) in one flex row. Only the evidence was
meant to give way (min-width: 0 + ellipsis), but the label kept the default
flex-shrink: 1. Flex shrinking is weighted by basis, so a command many times
wider than the label took almost none of the loss and the label was shrunk to
its min-content width - a single glyph for CJK, which may break between any
two characters. "没有未经验证的部分" rendered one glyph per line at every window
width.
The tick and the label now take flex: none; the evidence is the only item
that shrinks and ellipsizes.
The row also sat off the answer's column: it was indented by the tool-call
symbol column (42px), which top-level transcript items no longer draw, and
ran the full flow width instead of --doc. It now hangs like the answer card
(6px inline hang, max-width --doc), so the tick lines up with the answer
text and the activity group, and the row ends where they end.
Test: perf/cards.mjs feeds a turn_done receipt with a long && chain through
the real App in Chromium and measures the label at 1440/640/420px.
Before: "回执标签不被长命令挤成竖排 — 9 行" FAIL at all three widths.
After: 1 行 at all three, and the command is still clipped in-row.
The same page checks the row against the answer's text column: before,
tick off by 48px and row 60px wider at 1440px; after, 0px/0px at all three.
tsc --noEmit, vitest src/ui/cards src/styles (67 passed), make check clean.
Refs #10851
multi_edit returned on the first step whose old_string was missing, not
unique, or empty, so a batch with three bad steps took three round trips
to surface them all even though nothing was ever written.
The sweep now records a failed step, leaves the buffer as it was, and
checks the remaining steps against it. Any failure still returns before
the write, so the file stays untouched. A lone failure keeps its
existing "edit N: ..." form; several are listed under
"K of N edits failed; <path> left untouched", with the note that steps
after a failed one ran without its replacement, so a later failure may
be a consequence of an earlier one.
Test: TestMultiEditReportsEveryFailedStep (5 edits, 3 bad: not found,
not unique, empty old_string) fails on origin/studio, which reports only
edit 2, and passes with this change; the file is unchanged in both.
go test ./internal/tools/builtin/ passes; make check and golangci-lint
on the package are clean.
Refs #10792
The checkout under review is untrusted input, yet review built its grep
from the merged config and read the review skill with the checkout as a
project root. A reasonix.toml setting [tools.search] engine = "rg" and
rg_path = "tools/rg" made review's grep execute a program the checkout
ships, and a .reasonix/skills/review/SKILL.md without read-only replaced
the built-in review skill, handing an attacker-written prompt bash.
Review now loads the user-global config alone (Roots.LoadUserConfigReadOnly)
for its tool registry: search engine, rg_path and sandbox. Its skill store
has no project root, so project-scope skill directories are never
discovered. Provider and model resolution are unchanged.
On studio only the executor and the controller fired the session's hooks.
Every other agent that calls tools ran with no hook runner, so a PreToolUse
deny on reading a secret did not cover its reads:
- delegated children: TaskTool.subagentOptions (task, read_only_task,
parallel_tasks, fleet, background tasks and model-invoked run_skill via
RunProfileSpec) and roleWiring.skillRunOptions (read_only_skill) set no
Hooks. This ports main-v2's 896faf36d (#10843), which never reached studio;
- the dual-model planner (roleWiring.planner);
- the guardian, reachable through gateApprover -> guardianSess.Review;
- `reasonix review`, which builds its subagent outside boot.
hook.Runner gains ForRole(role): a runner firing the same live hook set in
the same environment, whose session id is the parent's id at fire time
suffixed with ":"+role. A SetSessionID rotation (/new, clear, branch
switch) therefore reaches every child, and a hook keyed on session_id
neither merges the planners of different sessions nor confuses a child
with the executor. The hooks and the repeat-suppression state live in a
shared hookSet, so SaveHooks/Replace reaches every derived runner.
boot passes the runner to roleWiring: planner fires as <session>:planner,
guardian as <session>:guardian (guardian.NewSession takes the hooks), task
and skill children as <session>:subagent:<ref|call id> through
TaskToolOptions.HooksForRole and skillRunOptions.
`reasonix review` loads only the user's hooks: hook.LoadOptions gains
Scopes, and boot.NewUserHookRunner loads hook.UserScopes (global
settings.json and installed plugins). Project hooks from the reviewed
checkout never run, since a branch under review is untrusted input. Each
review run fires under a fresh id.
Review hooks start in the checkout root so they can inspect it, which on
Windows would let cmd.exe resolve a legacy hook's bare `python` or `node`
to an executable the checkout ships, since cmd.exe searches the current
directory first. The review runner's spawner (hook.WithoutCwdExeSearch)
sets NoDefaultCurrentDirectoryInExePath=1 in every hook process, which
cmd.exe honours; the cwd stays the checkout.
Tests (real PreToolUse scripts that log their payload and exit 2):
- TestBuildRunsPreToolUseInsideTaskSubagent (boot.Build; task,
read_only_task, run_skill, read_only_skill): the child's read is blocked
and fires under <parent>:subagent:...
- TestEffectPreToolUseCoversPlannerReads / ...GuardianReads (boot.Build):
blocked, under <parent>:planner / <parent>:guardian
- TestEffectPlannerHookSessionFollowsRotation: NewSession moves the
planner's hook session to the new parent id
- TestReviewCommandRunsOnlyUserHooks: a global deny blocks the review's
read, a project hook in the checkout never runs, two runs fire under
two distinct ids
- TestUserHookRunnerDisablesCwdExeSearch: through the spawn seam, the
review runner's hook process starts in the checkout with
NoDefaultCurrentDirectoryInExePath=1
- TestLoadUserScopesSkipProjectHooks, TestRunnerForRoleDerivesSessionIDAtFireTime
Refs #9255
The 1.x line keeps the CLI's install channels until 2.x is stable, so
the npm and Homebrew job now runs only with the repository variable
STUDIO_PUBLISHES_CLI=true. Studio releases still carry the CLI archives
its remote workspaces install from.
A pull request authored with GITHUB_TOKEN never starts other workflows, so
the weekly acknowledgments PR never reported the checks branch protection
requires and could only land by bypassing it. The step now uses the
dedicated ACKNOWLEDGMENTS_TOKEN secret when it is set, and falls back to
GITHUB_TOKEN otherwise.
The contributor table also listed github-actions[bot]; bot accounts are
now filtered before the top 20 are taken (30 are fetched to keep 20).
Checked by running the script against the live API: the previous version
writes one github-actions row, this one writes none.
Wails stopped being the desktop shell when the Electron shell landed, but
seven places on `studio` still describe it in the present tense. Six of them
were already fixed on `main-v2` by #10018; they only need the same edit
carried over here.
- `CHECKPOINTS`: the desktop rewind control calls `controller.Rewind` over the
desktop host protocol, not the Wails binding.
- `TASK_CONTRACT`: the constraint keeps the desktop host JSON contract intact.
- `RECOVERY`: `doctor` inspects install problems without loading the desktop
shell, and macOS keeps LaunchServices launching the desktop app bundle.
- `DESKTOP_CRASH_DIAGNOSTICS_RUNBOOK`: the budget is pre-shell work, and
report/metric persistence runs after shell startup.
- `SESSION_MEMORY_RETRIEVAL`: the example project file no longer points at
`desktop/frontend/wailsjs/`, which went away with the Wails shell. Studio
commits its generated files under `internal/ext/extension/protocol/`.
- `REASONIX.md`: the frontends behind `control.Controller` are the chat TUI,
HTTP/SSE serve and the Electron desktop.
- `issue-auto-label`: the classifier prompt no longer tells the model the
desktop app is Wails.
Root cause: Electron 44 picks its Wayland Ozone backend by itself in a
Wayland session, and on drivers whose EGL platform init fails there (#10369)
the GPU process exits and the app dies before a window appears; the same
machine works with --ozone-platform=x11. Studio's deb entry runs the
Electron binary directly, so no wrapper builds its argv, and the only
switch main.js sets is disable-backgrounding-occluded-windows. Chromium
selects Ozone before main.js runs, so appendSwitch cannot choose it.
Decision: default to X11 whenever it is available, the same rule as the
1.x line. On Linux, when the session is Wayland (XDG_SESSION_TYPE, else
WAYLAND_DISPLAY) and DISPLAY is set, main.js relaunches once with
--ozone-platform=x11 prepended and exits, before it claims the instance
lock so the relaunch finds nothing held. REASONIX_OZONE_PLATFORM=wayland
keeps native Wayland, x11 forces X11, auto is the default. A pure Wayland
session without DISPLAY and an X11 session are unchanged, and an explicit
--ozone-platform or --ozone-platform-hint argument wins, which is also what
stops the relaunched process from relaunching again. The update handover's
app.relaunch() reuses process.argv and keeps the flag.
Test: two cases in test/unit.mjs. One loads the real main.js with a stub
electron module under a Wayland+DISPLAY environment and asserts it
relaunches with the flag and exits before requestSingleInstanceLock; on
origin/studio it fails with actual [['lock']] where
[['relaunch', ['--ozone-platform=x11', ...]], ['exit', 0]] was expected.
The other walks the session table through relaunchForOzonePlatform. Both
pass with the fix (38/38). A probe on Electron 44.1.0 confirmed a
top-level return after app.relaunch({args}) + app.exit(0) before ready
starts a child that receives the new argv.
Root cause: on DeepSeek thinking mode, reasoningOnlyFinishHonoured
accepts a finish_reason=stop with empty content as the final answer.
When that stop comes right after a tool round, the tool results never
get a visible synthesis and the turn goes idle until the user sends
another message (#9758).
Decision: a narrow structural retry. When this turn ran tools, no
visible assistant text follows the last tool result in the transcript,
and the stop is reasoning-only, the existing empty-final retry runs
once; a second reasoning-only stop is honoured. A reasoning-only answer
to a turn that ran no tools is still honoured, so the extra thinking
round the honour rule prevents does not come back there. The judgement
reads structure only: usedAnyTool, empty content, message position.
Non-DeepSeek providers keep their existing bounded retry.
Tests: TestRunRetriesReasoningOnlyStopAfterToolRoundOnce and
TestRunAcceptsSecondReasoningOnlyStopAfterToolRound drive a scripted
DeepSeek-policy provider through Agent.Run. On the base both fail with
"provider calls = 2, want tool round + reasoning-only stop + one retry";
with the fix they pass. TestRunAcceptsReasoningOnlyFinalWhenModelStopped
(no tools) passes unchanged, as do internal/runtime/agent and its
direct dependents.
read_file told the model to reference an image as @path and, if it
cannot read images, to delegate with read_only_task. Only a reference
in the user's own message attaches a picture, so for an image the user
never attached the delegation reached a child with nothing to look at.
The refusal now says an image attached this turn can go to
read_only_task, and that any other one needs the user to send it.
In 2,217 local sessions, all six image refusals were attachments, which
the delegation does carry; the unattached case has not been seen, so no
new task parameter is added for it.
build-host.js built reasonix-studio-host on Linux, and linux-payload.js
built reasonix-studio-update-helper, with cgo at its default. On the
ubuntu-22.04 runner that links both against glibc 2.35: in the
studio-v2.20.0 .deb both name /lib64/ld-linux-x86-64.so.2 and the host
needs GLIBC_2.32/2.34, so Studio will not start its kernel on glibc 2.31
systems even though the host is already held CGO-free by studio.yml. On
Linux the only cgo in its closure is the stdlib's optional net and os/user
glue, which have pure-Go implementations.
Both builds now pass CGO_ENABLED=0, and the package job reads the ELF
program headers of the .deb's Go binaries (packaging/elf.js) and fails if
one names a program interpreter.
Test: test/unit.mjs adds a case for elf.js; on the base it fails with
"Cannot find module '../packaging/elf.js'", with the fix 36/36 pass.
`node packaging/elf.js` on the studio-v2.20.0 .deb's host and helper
exits 1 ("... is dynamically linked (interpreter
/lib64/ld-linux-x86-64.so.2)"), and exits 0 on both built with
CGO_ENABLED=0 GOOS=linux.
Refs #10372
Root cause: legacy conhost with a CJK TrueType font (the default on
Chinese Windows) sizes East Asian Ambiguous runes by the glyph, so it
draws a font-dependent subset two cells wide: · … “ ” ‘ ’ — – ● → ↑ ↓
◆ ° × and Greek/Cyrillic letters among them. The TUI and its renderer
count all of them as one cell, so every full-width row carrying k of
them is k cells too wide on screen: conhost wraps it, the alt-screen
frame grows past the window and scrolls, and later repaints land on
shifted rows, which is the stacked-frame garbage in the report.
Decision: measure instead of guess. When stdout is a console, a
never-activated screen buffer of that console (same font) measures
each non-ASCII rune the layout counts as one cell, once, via
WriteConsoleW and a cursor read-back. A rune the console draws wider is
swapped, after layout, for a one-cell stand-in both sides agree on:
Unicode's halfwidth variant when it has one (→ becomes U+FFEB), else
the OS US-ASCII best-fit (· becomes ., “ becomes "), else the OS
default character. Each candidate is itself measured before it is
used. The swap is rune for rune at equal counted width, so applying it
to the finished frame changes no row's layout; escape sequences and
OSC payloads pass through untouched. A pseudoconsole (Windows Terminal,
VS Code, SSH) measures these runes as one cell, so nothing is replaced
there; other platforms get a nil fit and are unchanged.
The same measurement also covers what the inline mode prints above the
frame (answers and the banner), since a wrapped print adds rows the
renderer does not know about. The banner is printed from a command
goroutine, so the per-rune cache is guarded by a mutex.
Test: TestFrameFitsConhostMeasuredWidths renders the real TUI View at
120 columns after a streamed answer, with the width table measured on
a Chinese Windows 11 conhost, and asserts no row draws wider than the
terminal. With the substitution disabled it fails:
row 3 draws 121 cells on conhost, terminal has 120
With the fix it passes (also under -race). In a real conhost window
the unfixed binary shows the status row twice with "au" / "to" split
across rows; the fixed binary keeps the 30-row frame intact.
Refs #10540
The rendered config template writes bash = "enforce" even when nobody
chose it, so on a Linux host without bubblewrap the settings read back
as enforce and every later save was refused as sandbox.unavailable,
including one that only added a writable directory. Only switching the
jail on is refused now; a file already holding enforce saves its other
fields, and bash stays refused at run time as before.
Reproduced on a bwrap-less Linux: the new test fails there without this
change and passes with it. It skips where a sandbox exists and on
Windows, which forces bash off.
Root cause: the transcript was written only on the 30s autosave tick or
at turn end. A process killed between admitting the user message and the
next save reopened with the in-flight marker but without the prompt, so
the turn vanished from both the history and the model's context.
Decision: the same bounded change as the 1.x line's option A (1). The
agent gains an AdmissionSaver seam called from LandAuthoredUserMessage,
the one place a turn's user message lands, and the controller saves the
transcript there: one extra save per turn, skipped without a session
path, and a failed save only logs. The 1.x partial-output checkpoint
does not apply here: studio has no event-log session store, and the
cache-stable prompt prefix is untouched.
Test: TestKilledTurnKeepsItsPrompt (internal/session/control) runs one
turn, starts a second, streams part of a reply and ends the child
process with os.Exit (no TurnDone, no final save), then reloads the
session and runs recovery. On origin/studio it fails with "prompt of the
killed turn lost after restart"; with the fix it passes.
The test saves the sandbox settings twice. On a Linux runner without
bubblewrap the second save asks for a sandbox the host cannot run and
is refused, which has nothing to do with the switch under test.
When native mouse is on (REASONIX_DISABLE_MOUSE, or the /mouse toggle), the
terminal owns the mouse: the in-app scrollbar can no longer be dragged, so the
transcript drops its right-hand column and reclaims that width for content, and
fenced code drops its "│" gutter for clean copy. Capture-on is unchanged.
When --continue finds no saved session, it reports the miss and starts a fresh
session instead of exiting 1. Both the run path and the TUI resume path fall
through, a --continue miss is no longer reported as "continue" in telemetry,
and the flag help and docs say so.
The ask tool blocks the run on the user exactly like an approval prompt,
but only approval prompts fired the Notification hook, so an external
channel wired to it heard about every pending approval while a blocking
question looked like the agent still working. In YOLO mode ask is the
only prompt that waits on the user, and it was the one that stayed silent.
Fire the hook once the AskRequest is emitted, mirroring the approval call
site: message "answer needed: <first prompt>" (en/zh/zh-TW),
notificationType "question_prompt".
Rebuild of #9048 on studio.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Looked at on Windows Terminal against 1.x with the same scripted turn:
- A streaming answer is printed a block at a time, and each block lost the
blank line before it, so headings, lists, code and tables ran together.
A later stretch now starts with that line.
- Tabs were drawn zero cells wide in the full-screen view, so code lost its
indentation. Rows expand them before wrapping, as 1.x does.
- A list resumed after a code block restarted at 1; it counts from the
number it was written with.
- oneLine cut by runes, so a line with CJK text ran past the row and wrapped
to column 0. It cuts by cells.
- The approval panel and the end-of-turn mark drew ⏸ and ⚠, which Windows
Terminal renders as two-cell emoji over one-cell slots. The panel is headed
by its title in the accent colour, and its key hint is translated.
- The end-of-turn card is 1.x's: what went unproven, in the reader's
language, instead of "incomplete · 2 gap(s)".
- The footer's compaction figure reads "距压缩 81%" (headroom left), not
"压缩 81%".
write_file replaces the whole file, so a model writing from what it read
earlier silently discards anything changed since: an edit the user made
in their editor while the turn ran, a formatter, another process. The
run now keeps what the agent last saw of each file, through read_file or
its own writes, and write_file refuses when the file holds something
else, telling the model to read it again. edit_file and multi_edit match
their old text against the current file and were already safe.
In 2,217 local sessions, the case never shows in the logs, because an
outside edit is not recorded there, so this is a safety net rather than
a measured gain. It costs nothing until a file has actually changed,
and is on by default. [tools].protect_changed_files and a switch in the
Studio sandbox settings turn it off.
An explicit --model whose provider has no key failed with
`provider "deepseek-flash" requires DEEPSEEK_API_KEY`. Provider keys
resolve only from the loaded config's credentials file
(storedCredentialValue on Roots.UserCredentialsPath), so the message
sent users to export the variable in their shell or system environment,
which Reasonix never reads. It also labelled the model ref as a
provider. Reported through `reasonix acp` from the VS Code extension in
#10458.
The error now names the model, the variable, the credentials file of
the config it was resolved against, and `reasonix setup`, and says
shell environment variables are not read.
Test: TestResolveModelForCLIMissingKeyNamesCredentialFile sets the key
only in the process environment. On the base it fails with
error "provider \"deepseek-flash\" requires REASONIX_CLI_TEST_SHELL_ONLY_KEY"
does not mention ".../.reasonix/.env"
and passes with the fix; go test ./internal/frontend/cli/ passes.
GitHub acts on "Fixes #N" only for pull requests merged into the default
branch, which is studio, so fixes merged into main-v2 left their issues
open. On a merge into main-v2 this reads the description's closing
keywords, notes the fix on each open issue and closes it.
Adds [skills].suppress_warnings (default false), which quiets the non-fatal skill warnings printed during startup discovery. doctor diagnostics are unchanged. Implements #6483.
The terminal UI ignored REASONIX_DISABLE_MOUSE, which the guide still
documents. It now reads it as 1.x does: 1 starts with the mouse handed
to the terminal, 0 captures it, and unset leaves capture off over SSH,
where the native selection reaches the user's clipboard.
Co-authored-by: BuGlessRB <[email protected]>
A tag is a mutable pointer: whoever controls the action's repository can
move it to new code and every workflow here runs it on the next event,
with this repository's token and on its self-hosted runners. A commit SHA
cannot be moved.
Each reference keeps its tag as a trailing comment, the form Dependabot's
github-actions ecosystem reads and updates. This lets the repository-level
"require full-length SHA" policy be enabled without failing studio runs.
With no recovery_model or guardian_model configured, the Goal evaluator
runs on the main model. A DeepSeek thinking model can stream its whole
answer as reasoning_content and finish with stop and an empty content
block. boundedllm.Call collected only ChunkText, so the evaluator got ""
and every goal paused with "empty goal evaluator response".
boundedllm.Config gains ReasoningAnswer. When a caller opts in and the
provider reports finish_reason=stop with no content, Call returns the
reasoning stream instead. A truncated stream still yields nothing, and
reasoning past MaxOutputBytes is refused only when it would become the
answer, so reasoning before real content is never charged. The goal
evaluator opts in because parseVerdict extracts the JSON object and fails
closed on anything else; free-text callers (advisor, prompt refine) keep
the content-only behaviour.
Test: TestEvaluateReadsVerdictFromReasoningOnlyStop fails on the base
with "empty goal evaluator response" and passes with the fix;
TestEvaluateIgnoresReasoningOfTruncatedStream and TestCallReasoningAnswer
cover the stop guard, opt-in and output budget.
Refs #9678
resolveSessionQuery reported a query that matched several sessions as a
plain sentence, so neither caller could recover which sessions matched:
`reasonix run --resume <q>` exited 1 with only a match count, and the
terminal UI refused to start the same way.
The ambiguity is now a typed ambiguousSessionQueryError carrying the
matched sessions. The headless start still exits 1 but lists each
candidate (last activity, the id --resume accepts, its title) followed by
a new AmbiguousResumeHint. The terminal UI, which already requires a TTY,
opens its session picker limited to those matches (tui.Options.PickAmong).
The narrowing belongs to that first picker alone: it is dropped whether
the first session fetch succeeds or fails, so a later /resume lists
everything.
The picker resumes the chosen session in place, so an ambiguous query
with --copy is refused up front with the same candidate listing rather
than opening a picker that would silently ignore --copy.
Tests: TestRunAmbiguousResumeListsCandidates drives runAgent with an
ambiguous --resume and fails on the base with only
`session query "session" is ambiguous (2 matches)` on stderr;
TestTUIAmbiguousResumeWithCopyIsRefused fails when the picker path is
taken with --copy set; TestPickerNarrowingEndsWithAFailedFirstFetch fails
when the narrowing is cleared only after a successful fetch;
TestResolveSessionQueryAmbiguityCarriesMatches and
TestPickerStartsOnTheGivenSessions cover the typed error and the
narrowed picker.
recall() kept no draft slot: up on a one-line composer overwrote the
text with the previous prompt, and down past the newest entry called
composer.Reset(). With any history present, a single down on a draft
that had never left for the history wiped it outright.
The draft is now set aside when a walk begins and restored when the
walk returns past the newest entry; down at the live position is left
to the textarea instead of resetting it.
Tests in internal/frontend/tui/recall_test.go: TestRecallReturnsTheDraftItLeft
and TestDownOnADraftKeepsIt fail on the base (the composer ends empty)
and pass with the fix.
Refs #10425
A CLI process that died from a panic on a background goroutine or a
fatal runtime error (out of memory, concurrent map writes) left no
evidence. runWithCrashCapture only recovers panics on the main
goroutine and nothing called debug.SetCrashOutput, so that output went
only to the console's stderr, which is gone once the terminal closes.
The CLI now mirrors runtime crash output into a per-PID file under the
Reasonix home. The next start turns any dump left by a dead process
into a pending crash report that keeps the runtime classification and
the sanitized goroutine stacks and redacts the panic value, and removes
the empty files of processes that exited without one. A panic already
reported through recover detaches the file first so it is not queued
twice.
Test: TestGoroutinePanicIsReportedOnNextStart re-runs the test binary
as a child that panics on a background goroutine, then asserts the next
start lists exactly one report with the panicking goroutine's stack and
no panic value. On the base it lists none.
Dispatch-only, so nothing a pull request writes runs on that machine.
The version is a prerelease one patch above the newest release, so the
build neither offers itself that release nor hides the next one.
A comment on a pull request needs pull-requests: write. A refusal that
is not a rate limit is final, so it fails the entry at once instead of
being retried for ten minutes.
A 1.x command line keeps working under 2.x: workspace-write is Auto,
danger-full-access is Yolo, and read-only, which 2.x has no mode for,
falls back to Ask. MIGRATING says where yolo's meaning differs.
Co-authored-by: BuGlessRB <[email protected]>
In 1.x a bare reasonix, or one given only session options, was the chat;
in 2.x it printed usage, so -c, -r and every muscle-memory command line
stopped at a help screen. In a terminal it now starts the terminal UI;
without one it still prints usage, since no one is there to chat with.
The terminal UI takes the rest of 1.x's session options: --effort,
--permission-mode (plan starts in plan mode, acceptEdits allows the write
tools), --yolo and --dangerously-skip-permissions, --add-dir,
--allowed-tools, --copy, --max-steps and the retired --profile. --yolo and
--dangerously-skip-permissions were listed as session options but nothing
read them. The usage text names the interactive form again.
The migration guide says 1.x 1.38.8 and later conversations now open in
2.x, where they are imported, how an import follows 1.x, and that the
machine session list shows them once imported. The roadmap marks W1b done,
which meets gate G2, and points W2 at its pull request.
1.x 1.38.8 and later keep every conversation in <project>/sessions-v4, a
directory per session holding a manifest and an events.frames log of
checksummed batches of zstd-compressed JSON, with large payloads and images
in a content pool beside it. 2.x did not list them, so a person moving from
npm 1.x found their recent history gone.
internal/state/sessionv4 reads that store and nothing else: it validates
each batch's framing and checksum, treats a batch without its end record as
a write in progress, resolves pool references by digest, and folds the
message events into the conversation 1.x shows, with origin "host" read as
HostAuthored and pooled images carried as data URLs. It takes no lock and
writes no byte under the root.
Session lists now include each v4 conversation that has no import yet, at
sessions/v4-<id>.jsonl, using 1.x's own listing cache when it still matches
the log. Opening that path imports the conversation there first, recording
its source; opening it again after 1.x continued it refreshes the copy only
while 2.x has not added to it. The name carries a prefix because 1.x takes
sessions-v4/<name> as the mirror of sessions/<name>.jsonl. Verified end to
end against stores written by 1.x 1.38.10 and main-v2 HEAD, on macOS and
Windows.
A reply batch needs one small job, and on GitHub-hosted runners it queued
behind every branch's CI matrix. The win-gpu runner takes only jobs that
name its label, so replies start at once and CI keeps its runners.
A batch of triage replies from a maintainer's own account runs into
GitHub's content-creation limit, which answers "submitted too quickly"
and stays closed for a while. This manual workflow takes a JSON list of
replies, posts them as github-actions[bot] under the repository's quota
at a steady pace, retries a refused comment, skips a reply already there
and closes the issues marked for it.
The migration guide said 1.x writes schema 2 from 1.38.2 on. It does so
only through 1.38.7; 1.38.8 and later keep conversations in sessions-v4,
which 2.x does not list. The guide now says where each 1.x version keeps a
conversation and which of them 2.x opens, drops the REASONIX_SESSION_LOG
step that 2.x reading schema 2 made unnecessary, and states that 2.x never
writes a 1.x session's files.
The roadmap records the direction the version announcement left open: the
gates that end 2.x pre-release in the same release that moves npm and
Homebrew, the planned weekly release train and update channels, the 1.x
feature and format freeze with maintenance ending 3 months after general
availability or at a backstop date, and splits the session work into the
done schema 2 half and the open sessions-v4 half.
1.x 1.38.2 through 1.38.7 save sessions in a schema 2 event log, an
append-only DAG of messages under named heads, and 2.x refused every one of
them with "uses schema 2; this build supports up to 1". 2.x now replays
that log read-only and materializes the head 1.x would open: parent chain,
patches, redactions and system override applied, and 1.x's origin "host"
read as HostAuthored so injected context stays out of the transcript and
the preview.
2.x never writes the log or anything beside it. A save that finds the
transcript unchanged stops with ErrSessionLogUnchanged, so opening a 1.x
conversation touches no file. Before a turn starts, and at any save that
has more than 1.x holds, the controller moves the conversation to a new
2.x session with the same title and says so; 1.x keeps the original byte
for byte. Verified against sessions written by a real 1.38.7 build.
Host-authored messages no longer count as a turn or become the preview,
and the event-log probe moves to its own file.
docs/MIGRATING.md was the 0.x to 1.0 guide and told a reader on the 2.x
default branch that main-v2 was the active line. It now covers moving from
1.x to 2.x: what the two lines share under one Reasonix home, which sessions
open where now that 1.x 1.38.2 writes a schema 2 log 2.x cannot read, the
REASONIX_SESSION_LOG=v1 switch that keeps a machine running both usable,
the commands that moved, and the steps to run 2.x beside 1.x.
docs/ROADMAP.md is the roadmap the version announcement (#10748) promised:
how 2.x ships, the terminal takeover that gates the release that moves npm
and Homebrew to 2.x, the remaining workstreams, and the decisions still
open. Both are owned by @esengine, and README links to them. The guide's
old prose debt leaves the repolint baseline.
1.x and 2.x share one session directory, and 1.x writes <id>.turns.jsonl
beside each transcript. 2.x classified it by its .jsonl ending and listed
it as a conversation of its own, so a machine that ran both showed a ghost
session for every 1.x conversation.
An Anthropic `error` event (overloaded_error and the like) that arrives
before anything was emitted left the provider as a plain
StreamPayloadError, so the Agent never replayed the attempt, while the
OpenAI provider already promotes the same pre-output payload error to
StreamInterrupt(upstream_error) (#10719). Apply that rule here: the read
loop's sink records whether a chunk reached the consumer, and only an
error with nothing emitted becomes an interruption. After output a
replay would show it twice, so it stays a payload error; the payload
identity survives the wrap for errors.As.
sendChunk moves into the sink so anthropic.go does not grow past its
size budget.
A disk catalog runs in WAL mode with no journal_size_limit, so after a
burst of writes the -wal file keeps its high-water size for as long as
the database is open (#10714). The limit is a per-connection pragma and
the pool opens up to MaxOpenConns connections, so it rides the DSN the
way busy_timeout and foreign_keys do rather than a single Exec that
would reach only one of them. 4 MiB leaves ordinary checkpoints alone.
Relays that re-serialize tool schemas emit `required: null` for an
object schema that omits "required", and the upstream validator rejects
the whole request with "null is not of type 'array'" (#10614).
context_budget is one such schema on this branch, and any MCP or plugin
tool may be another, so the fix sits in canonicalization rather than in
each tool: a root object schema gets an empty array when it has none,
including after an invalid non-array value is dropped.
This changes the tool block of the cache-stable prefix once (golden
ToolsHash/PrefixHash regenerated); the budgets rise by the bytes the
array costs: provider-visible 3700 -> 3708 tokens, built-in 6767 -> 6787.
Ctrl+V (Alt+V on Windows, whose terminals keep Ctrl+V for their own text
paste) reads the clipboard of the machine the terminal runs on. An image
goes to the kernel through /attachments and stands in the composer as
[image #N], sent as the @-reference it was stored under; without one the
clipboard's text comes in the way a terminal paste would, which is also
how a console without bracketed paste gets a paste folded.
internal/platform/clipimage reads the image through what each platform
ships: osascript, Windows PowerShell, wl-paste or xclip. A tool that cannot
say what the clipboard holds counts as offering no image rather than being
told apart by the words of its error, since either way the paste falls back
to text.
The inline renderer places a print by scrolling it in above a frame it
takes to sit at the bottom of the screen, and places each print from where
its last redraw left the frame. A restored conversation arrived as dozens
of prints with no redraw between them, on a screen the frame did not yet
fill, and landed interleaved with lines lost. A commit now goes out as one
print, split into pieces that fit above the frame with a redraw between
them, and a replay first fills the screen so the frame starts at the bottom.
The cancelled-turn, compaction, shell-busy, stream-reload and declined-
question lines now come from the i18n catalog.
Full screen, a shell call's output shows its ten-line preview with a
"more lines (Ctrl+B)" row, as in 1.x: Ctrl+B opens the newest one up to
200 lines and shuts it again, and a click on that row does the same. A
settled block keeps the row it draws, so opening it redraws only that
block. A drag held against the viewport's top or bottom row keeps
scrolling the selection until it leaves the edge or the transcript ends.
A bare -r opens the saved-session picker on the first frame, and /resume
opens it mid-chat, as in 1.x: typing filters, the open session is marked
and skipped by default, and choosing one binds the runtime to it through
/resume and reads it back. /resume and /mouse are offered in the slash
menu beside the kernel's commands. /sessions dates each entry, and moves
to its own file.
The TUI's event sink now carries the configured system notifications, as
the 1.x chat screen's did.
The transcript is drawn through its own viewport, as 1.x drew it: a
scrollbar in the last column, the wheel, PgUp/PgDn and Ctrl+Home/End,
dragging the thumb, and drag-to-select that copies on release. Settled rows
keep how to draw them, so a resize rewraps them. /mouse hands the mouse
back to the terminal for its own selection, and --inline writes the
conversation into the terminal's scrollback instead.
Users coming from 1.x meet the screen they know: the banner and tip, an
accent-framed composer with the ❯ prompt, the mode tag on its colored
ground, the footer's model/effort group and its cache, context, fold,
balance and cost row, a "本轮" usage line under each request, the ▎
thought marker, the ⎿ tool summaries, the To-dos panel, and approvals and
questions as selectable panels in place of the composer. Labels come from
the shared i18n catalog, so the screen speaks the configured language.
Shift+Tab steps ask, auto and plan; Ctrl+Y toggles YOLO.
The terminal UI reads as a product rather than a log: a welcome card names
the model and workspace, the composer sits between rules with a placeholder,
a running turn shows what it is doing and for how long, the status bar
carries a mode badge and a context gauge, approvals and questions are boxed
with their keys standing out, and the user's own lines are a band.
The theme is now read from config before the program starts; it was left at
the built-in default.
A message frame that arrives after the call which closed its answer now
settles that answer instead of appending a second copy, which printed every
answer before a call twice.
The page listed every model as a card to click for the primary model,
then drew the per-task roles as a second set of dropdown cards below it,
so choosing what runs where meant reading two layouts that said the same
kind of thing. The 1.x desktop app put this in one table, and #10761 asks
for its layout back.
Each use is now a row: what it is for, a select of the models that can do
it grouped by service, and the service that model goes through. The
default model's row also lists what that model declares: image input,
adjustable reasoning, context size. Roles that follow the default say so, and
decision says it has no source rather than offering chat models. The two
settings rows it replaces become one, 按用途选择模型; reasoning effort
and context upkeep stay below it.
The card list and the role picker are gone with their styles. The
services page now works out its default selection while rendering, so
it no longer flashes an empty detail before the selection lands.
Settings kept every model setting on one long page: the source list with
an editor that unfolded inside it, the primary-model list below that, and
roles and context after. Adding or changing a service meant finding the
right row, opening it, and scrolling past the rest. The 1.x desktop app
split this into 模型服务 and 模型偏好 with a list-and-detail layout, and
users asked for that back (#10761).
模型服务 is now its own section. Services are listed on the left, with a
search once there are six or more and 添加模型服务 at the foot; the
selected service's fields are open on the right without a second click,
and adding a service takes the right side while the list stays in
place. The page opens on the service in use, or on the one the composer
sent the user to declare effort levels for. At narrow widths the list
stacks above the detail.
The remaining section is renamed 模型偏好 and keeps the primary model,
reasoning effort, per-task models and context upkeep. The composer's
manage-models entry, its effort-declare link and the palette entry now
open 模型服务.
The composer's permission menu and Settings each kept their own table of
the four approval modes, and they drifted: the menu called ask 逐项确认
and auto 自动继续, Settings 询问 and 自动, the shelf 自动批准, and the two
lists ran in different orders. Users read the menu's 逐项确认 as a mode
Settings did not have.
approvals.ts is now the one table. Menu, Settings and shelf all say
不询问 / 询问 / 自动批准 / 全部放行, ordered from strictest to most
permissive, with the same descriptions.
In a narrow conversation pane the menu opened rightward from its trigger
and ran off the screen. There it now spans the toolbar row instead.
- Questions: an open ask is answered one question at a time. A digit
answers a single choice or toggles a multi choice, typed text joins the
picks as an answer no option offered, Esc declines with nothing
selected. The card left in the scrollback says what was answered.
- Completion: `/` at the start of a line and an `@` token open the
kernel's menu as they are typed, Tab asks anywhere. Offsets cross
/complete in UTF-16 both ways, so a line with CJK before the token
splices where the kernel meant; a directory keeps the menu open a level
down.
- The task list is re-read from /todos when the kernel says it moved and
drawn while it has work left; a finished list goes away.
Found running against a real model: a thought-only step and whitespace-only
answers printed an empty speaker header; option and menu rows lost their
indent to the row-shortening helper; the receipt read the verdict "done"
as unfinished.
A `!` command ran and was shown, and the model never saw it: the next turn
had no idea what the user had just looked at. RunShell now follows the
command with a turn whose input is the command, its exit code, any launch
error and its output, so the model responds to it the way it would to a
line typed to it.
The output is bounded at 24 KiB, cut from the middle with the cut said, so
a large listing neither fills the context nor reaches the model looking
complete. A command the user stopped is not followed by a turn, and neither
is one run where no model is attached. The ACP `!` prompt now ends after
that answer.
Expanding a folder in the explorer walked the whole workspace and kept
the entries whose parent matched, so one folder the process could not
read anywhere in the tree failed every expansion with a 500, and a large
tree made every click slow; the root listing, which skips into nothing,
kept working, which is why the list drew and its folders never opened.
The explorer drew that error in the file view's slot, which the docked
file list hides, so the click looked like it did nothing.
A folder listing now reads that folder alone, and its real path is held
inside the workspace so a link cannot list what is outside it. A search
still walks the tree and passes over a folder it cannot read. A missing
folder answers 404 workspace.file_missing, a linked-out one 400.
The explorer now shows its own failures in the file list. Reloading the
tree, which happens when the changed files change after a turn, reads
the folders that were open again, where it used to fold them all shut.
Checked on Windows against a workspace with a folder denied by ACL: the
folders beside it open, the denied one reports in the list, and an open
folder stays open through a reload.
The second layer of the terminal UI: the program a person operates, on
top of the client, stream and transcript.
- `reasonix tui [--model] [--preset] [--dir] [-c] [-r QUERY] [prompt]`
assembles the kernel the way `reasonix serve` does, adopts it into a hub
and drives it through Hub.InProcessClient: no port is opened. The kernel's
logging goes to ~/.reasonix/logs/tui.log while the UI owns the screen.
- Rendering keeps the terminal's scrollback: every settled row goes above
the program in the order it happened, and a streaming answer's finished
Markdown blocks go as soon as they are final (nothing inside an open
fence). What is still changing stays in a live region capped at half the
screen, each row cut short of the terminal width so none wraps into a row
the renderer does not know it drew.
- Input: Enter sends, or queues while a turn runs; Ctrl+S steers; Ctrl+J
(or Shift+Enter where the terminal reports it) is a newline; Esc
interrupts; `!` on an empty line is shell mode; Shift+Tab cycles the
approval mode; Ctrl+C twice or Ctrl+D on an empty line quits. Large pastes
fold into a token and go to the model whole. Approvals take only the
answers the host says it honours; plan approvals map to start, revise or
leave.
- The transcript seats a turn on the message it began on, the way Studio
does: a queued follow-up moves to where its turn started, steered input
waits for its steer event, and another client's turn is drawn once from
the kernel's text. Restore rebuilds from /history for -c/-r and after a
stream gap.
cli.go's argv normalisation moves into normalizeCommand so the dispatch
switch keeps one case per command within its complexity budget, and `tui`
joins the shell-completion registry.
The theme, colour profile, Markdown/math renderer, tool cards, diff view,
assistant block and clipboard writes move out of internal/frontend/cli into
internal/frontend/termrender, so a terminal frontend other than cli can draw
the same bytes without importing cli. cli now calls termrender for every
style, width, box and Markdown helper; view_helpers.go stays in cli.
termrender exports the surface a frontend draws with: Bold/Dim/Green/Red/
Yellow/Accent/Reverse, VisibleWidth/PadRight/Boxed, ConfigureTheme,
ConfigureThemeFromConfig(ForTTYOutput), ActiveTheme/ThemeName and the
ThemeStyle/ThemeFg/ThemeLipColor/ApplyTextareaTheme helpers over an exported
Palette/Color, RenderMarkdown and MarkdownRenderer, AssistantBlock, ToolCard/
ToolDot/ToolHead/ToolArg/ToolDisplayName, DiffBlock/DiffStat, CopyToClipboard
with ClipboardCopyMsg, ANSIConsoleReady and SetColorProfile.
Transcript helpers nothing calls (edgeScrollDir, selSpan, scrollbarThumb)
are dropped with their tests, and the clipboard message loses the seq and
status-hint fields no caller set. The metadata-command test now asserts that
version/help never reach the TTY theme setup, through a cli-side seam, since
the terminal probe it used to swap is termrender's.
The repolint baseline diff is the moved files' prior essay debt carried to
their new paths (mathnode.go, md.go, width_test.go; transcript.go's is paid
off), plus orphan budget for the eight exports whose consumer, the new terminal
UI, is not in the tree yet (ToolCard, DiffBlock, AssistantBlock,
RenderMarkdown, CopyToClipboard, ApplyTextareaTheme, ThemeName) or which only
a cli test drives (SetColorProfile). Wiring the UI reclaims that budget.
normalizeNullRedirects treated a backtick as an escape inside any quote,
so in echo 'a `' >nul the backtick swallowed the closing single quote,
the string never ended, and the real >nul after it was left alone; under
bash that writes a file named nul. Single-quoted text is literal in both
bash and PowerShell, and the backtick is PowerShell's escape only inside
double quotes, which is where the backslash case already applied.
The new cases also cover the escaped-quote branches, which had no test:
the sandbox package fell to 77.6% against its 78.1% coverage floor on
the lint job, three statements short.
fc3a44dbc moved jobs from internal/runtime to internal/tools, and the
pull-request race step still listed ./internal/runtime/jobs/..., so every
pull request's race job failed with setup failed on a directory that no
longer exists. Pushes run the full ./... sweep and never saw it. Every
other path the workflows and the Makefile name still exists.
* fix(pricing): bill Chinese public holidays as DeepSeek off-peak
Port of the main-v2 holiday fix (#10756) to the studio pricing package,
which moved from internal/billing to internal/contract/pricing and now
carries the peak schedule in PeakWindow. The holiday table rides the
same struct field pattern as WeekendOffPeak: PeakWindow.HolidayOffPeak
lists Beijing-local dates off the State Council 2026 notice, and
IsPeak flattens them like weekends. Make-up workdays (调休) stay
off-peak: they are calendar weekends, which is how DeepSeek rules them.
* chore(pricing): keep the holiday comments within the repo's comment rules
English only and within the essay budget; no behaviour change.
---------
Co-authored-by: nanami-0713 <[email protected]>
Co-authored-by: yhh <[email protected]>
The first layer of the new terminal UI, which drives a session through the
same HTTP/SSE contract Studio uses and imports nothing from the controller:
- Client: typed calls for the routes a chat needs; a refusal carries the
kernel's code (APIError), never only its sentence.
- Subscribe: /events with Studio's ordering — numbered frames delivered once
and in order, a jump filled from /events/replay before the frame that
revealed it, what the replay no longer holds reported as a gap, a
backwards count read as a restarted stream, a reconnect resumed with
Last-Event-ID.
- Transcript: the fold of Studio's session reducer — the settled message
repairs the streamed answer, one card per call with sub-agent calls under
their task, prompts settled here or by the kernel's receipt, pending input
moved to where the turn read it, and turn_done saying which of four ways
the turn ended.
An end-to-end test drives a real kernel behind an in-process hub through a
turn that streams, waits on an approval answered through the client, runs a
command and ends settled.
The editor is a frontend in the kernel's process, like the terminal UI
the in-process serve client exists for. A prompt that starts with `!` now
runs through RunShell instead of going to the model: it streams back as a
tool call and ends the prompt when the command finishes; cancelling the
prompt cancels the command.
Hub.InProcessClient reaches the hub's routes with no listener: a frontend
in the kernel's own process opens no port, so there is nothing on the
machine or the network to authenticate against. Responses stream through a
pipe as the handler flushes, so /events works through it.
A `!` line on /submit is the user's own shell command, which runs
unconfined. It stays refused off a socket (shell.unavailable_over_http);
through the in-process transport it runs. The transport marks the request
context itself, and no request read from a socket passes through it, so the
grant rests on which way the request came in rather than on anything it
carries.
Every ended job drew the same green pip and printed the kernel's raw status
word. Only done is green; failed is red, killed and interrupted are neutral,
and each is labelled in the reader's language.
The background-jobs deck listed jobs with no way to stop one: the controller
had CancelJob, but no port carried it and serve had no route. BackgroundJobs
joins SessionAPI, POST /jobs/{id}/cancel drives it and refuses a job that is no
longer running as job.not_running, and each running row gets a stop button.
On touch the deck could not be closed: it opened on :hover and :focus-within,
which a tap leaves set, so the chip's own toggle had no effect. Hover opening
is now limited to hover-capable pointers, and a press outside or Escape closes
it; the context card gets the same treatment.
The meter rail did not fit a phone: the chip ran past the edge, and tapping it
scrolled the overflow-hidden .app sideways, dragging the composer with it.
Narrow panes drop the sparkline, wallet, speed and (narrowest) cost readings,
the two popovers position against the rail instead of their chips, and .app
clips rather than hides so focus can no longer scroll it.
perf/composer.mjs counts only visible peers as overlapping the model button.
bypassed() matched only exact entries and a leading "*". The bypass list
Clash, v2rayN and xray write by default is `<local>;localhost;127.*;10.*;
...;192.168.*`, so every trailing-wildcard entry missed. It also lacked
WinINET's implicit rule that loopback goes direct unless `<-loopback>` is
listed.
In the default auto mode, on Windows with such a system proxy, requests to
local and LAN model endpoints went through the proxy: Ollama or LM Studio
on 127.0.0.1, vLLM on 192.168.x. That added a hop, failed outright when the
proxy routed to a remote exit, and made a cancelled stream keep the local
model generating. Cancel closed only the client-to-proxy hop, and the proxy
held its upstream connection open.
"*" is now a wildcard anywhere in an entry. Only "*" is special, and
brackets are trimmed from IPv6 entries. localhost, 127/8 and ::1 bypass
unless `<-loopback>` is present. TestBypassed gains the trailing-wildcard,
near-miss, mid-pattern and loopback cases; it fails on 8 of them without
the fix. The same change lands on main-v2.
The CLI had no install channel on the 2.x line: release-studio.yml builds the
CLI archives for Studio's remote installs, and npm's `reasonix` and the
Homebrew tap were published only by the retired 1.x release machinery. The
2.x line now takes both over, and 1.x no longer publishes the CLI.
A cli-channels job runs after the release is published:
- npm: the existing npm/build.mjs builds the per-platform packages and the
launcher and publishes them through publish.mjs's staged dist-tag flow. A
2.x version outranks every 1.x one, so `latest` moves to it; a candidate
goes to `next`.
- Homebrew: scripts/render-brew-cask.sh writes the cask from the digests in
the release's own SHA256SUMS — a missing digest fails the render rather
than shipping an empty one — and the job pushes it to
esengine/homebrew-reasonix. Candidates leave the tap alone, since brew has
no prerelease channel.
The Studio release carries the CLI archive too, listed first. Double-clicked
from Explorer, bare argv printed the usage and exited, and the console window
Windows opened for it closed with it, which read as a crash. When this
process is the only one attached to its console, it now names Studio's
installer and waits for Enter; from a shell nothing changes.
The usage examples dropped bare `reasonix`, `--continue` and `--resume`, which
only print the usage on this line.
LAYOUT.md: three list items split to stay under the prose width.
Every workspace holding a warm runtime drew a raised box, so four rows read as
four selections and people asked how to close them; the kernel already retires
idle panes at capacity. Only the workspace of the active session keeps its
emphasis, and a session held by a pane keeps its pip.
The project row's × removed the project from the list but sat next to + where
it read as "close". It is now a ··· menu (workspace.menu) whose one item leads
to the existing confirm.
The coverage job installs no bubblewrap, so an enforced bash sandbox refused
the background job and the test read no job at all. It measures how a job's
end is attributed, not where it runs.
Product:
- Tool cards wrap their head in details.tool-disclosure, so the hit highlight
and the failed-call peek never matched them; both selectors now reach it.
- Light --faint fell below AA on every raised ground and overrode the tiered
tokens; darkened by lightness only. Segments and the new-task shortcut sit on
--overlay and read --muted. Hard-coded crumb separator, delta and failed
state colours now use tokens; the approval tool tag uses --warn-ink, the
running step reads a darker --net, and the failure pill mixes its ink
towards --text so it clears AA in both schemes.
- The local workspace row's remove control is the same Cross icon the remote
rows already draw, not a text glyph.
Guards: retargeted to the current markup (model picker trigger, tool
disclosure head, context card MCP rows, wallpaper section, queue and plan
controls), dropped checks for the removed task board and focus mode, and the
wallpaper reach is measured with the dim overlay cleared. All 22 exit 0 and
selfcheck shows each can still go red.
Run — the synchronous turn `reasonix run` drives — never took the turn gate
the guarded paths check. A turn admitted through the gate, such as the
inbox delivering a background job's end, could therefore start on the same
executor while Run was mid-turn, and the two wrote the task state together:
the race TestEffectRestartNamesWhyABackgroundJobEnded tripped on CI.
Run now admits through runSynchronousTurn, as RunTurn does, and waits out a
turn already in flight rather than failing: after a resume the host may
dispatch a job's interruption before the caller's first Run. Its turn
completion is recorded once, by the admission, instead of twice. Cancel()
now reaches a headless turn as it does any other.
The restart test's scripted model started a new job on any turn it did not
recognise, which the interruption delivery now visibly is; it only
acknowledges such turns.
Run against studio's own stylesheet, the browser guards reached parts of the
window they had never measured.
- A collapsed column is inert, not only narrow. The rail was inert only in
focus mode and the workbench never, so Tab walked into controls nobody
could see and scrolled a column of zero width to show them. Both are inert
while collapsed, and the workbench stops painting once its collapse ends.
- Focus mode is gone. Its only way in was removed with the side column in
the studio redesign; the state, the Escape branch, the action id and the
layout notes that described it are removed with it.
- The queue's withdraw control lost its enlarged hit area when it moved from
the transcript into the queue panel: 19px tall. It gets the outward
::after back, 29px.
- Switching to the run analysis gave the transcript and the docked workbench
the same view-transition name, so the browser aborted the transition and
the rejection went uncaught. The docked workbench no longer takes the
name; it keeps it when it is the tab being switched to. A transition the
browser skips is not an error.
- The studio composer's own chrome is taller than the one the queue's
shared ceiling was set against, so 64 queued lines left the transcript at
36-39% of the pane. Studio lowers the ceiling (28vh, 19vh when short) and
the transcript stays above 40%.
The egress grant test sat between TestEverySubjectSensitiveToolActuallyAsksSomething
and its comment, which read as one six-line block. Windows has no /tmp and no
sandbox that reaches the proxy on a socket.
On Linux the list only warned: bubblewrap either kept the command's network
namespace (no network at all) or dropped it (all of it), and a separate
namespace cannot reach a proxy on the host's loopback.
The command now keeps its namespace, which has no route out. The egress
proxy also listens on a Unix socket under the user's cache, mounted
read-only into the sandbox, and the host binary runs inside the namespace
as the egress bridge: it listens on the namespace's 127.0.0.1:<proxy port>,
carries each connection to the socket, runs the command, and returns its
exit status as a shell would. HTTP_PROXY and friends need no change. The
bridge is an entry of whatever binary imports the sandbox package, chosen
by an environment variable it clears before the command starts, so no
helper has to be shipped or found.
The host's own loopback is out of reach from the namespace, so a local
proxy cannot be used to go around the list. A route with no socket keeps
the namespace shut rather than falling back to open.
The bridge's forwarding and exit status are tested on any Unix; the
namespace boundary and the end-to-end refusal and approval run where
bubblewrap is usable (Linux CI).
A host the egress list does not name was refused outright. Where the
frontend can ask, the proxy now holds the connection and asks through a
network_egress approval whose subject is the host: allow once (for that
command), allow for the session (that host only), or decline. A declined
host reaches the model as declined-by-user rather than blocked-by-allowlist.
- network_egress is declared subject-sensitive in permission, so a person
answers whatever the approval mode, an unattended session refuses, and a
session grant names the host it covers. Before that declaration a grant
for one host answered for every host after it.
- The controller's sandbox approver also answers ApproveEgress; bash finds
it on the tool context the escape approver already rides, so nothing
below the controller grew a new channel. Headless runs have none and keep
refusing.
- Each command asks once per host, one question at a time, and a token's
asker ends with the command. Denied hosts are never asked.
The guard asserted that the border glow runs while a turn runs; studio's
design removed that glow, and its running signal is the run mark's R being
traced (studio-r-run). The three directions stay: still at rest, moving
while running, still again after.
It also flagged the mark's straight stroke as drawn on nothing, because an
SVG shape's box leaves out its stroke and a straight line has no area. An
SVG shape is now judged by its drawing's box.
In the light theme the new-session button and the session filter segments
were separated from the sidebar by fill alone, a 1.8 step in OKLab
lightness, and the button had its border removed: neither read as a layer.
Both now mix 3% of the text colour into their fill, in light only.
Each session row reserved 34px of its title for the menu button and drew
the button at zero opacity until hover, so every title was cut short for a
control nobody could see. It now rests at 35% and comes up on hover, focus
and selection as before.
In a conversation narrower than 460px the ask card's actions stack, and they
stacked in DOM order: "reply without choosing" above the primary answer.
The approval card already leads with its main action there, and the cards
guard holds both; the primary button now takes the first row.
The refine button is disabled until there is a draft, yet its icon still
showed a pointer and it lifted on hover, reading as something to press.
A disabled mode button keeps its tooltip but takes the default cursor and
does not lift.
Network in the bash sandbox was one bit: open to everything, or shut. With
[sandbox] allowed_domains set, bash's external traffic goes through a
host-side proxy (internal/safety/egress) that admits only the listed hosts;
denied_domains outrank them.
- The proxy serves CONNECT and absolute-URI HTTP on loopback, decides from
the requested name, refuses names that resolve only to loopback,
link-local, metadata or this host's own addresses, and chains through the
upstream HTTP or SOCKS5 proxy the environment names.
- macOS: the Seatbelt profile shuts every destination but loopback, Unix
sockets and the proxy port, and closes the loopback ports of the proxies
the environment and the system proxy settings name. Loopback stays open
so tests and dev servers keep working. Bind, inbound and outbound are
separate grants, because (local ip "localhost:*") on network* matches the
local end of every outbound connection and admits all of them.
- Each confined launch carries a token as proxy userinfo; what the proxy
refused under it is appended to the bash result by the host, so a policy
refusal reaches the model as one instead of a bare connection error.
- The list only narrows open network: it applies where network is on and
the sandbox confines; a repo may set one only where the user has not, and
its denials only add to the user's.
- Linux and Windows warn that the list is not enforced there.
Known gaps, stated where they live: a host proxy on a loopback port nobody
declared stays reachable, name lookups still reach the system resolver,
the decision reads the requested name (domain fronting is not caught), and
MCP servers are not routed through the proxy.
internal/safety/egress is declared sensitive, with its coverage floor.
The window's error bar sat in the main column under the panes: every failed
request pushed the composer up by the bar's height, as a full-width strip at
the far end of the screen from most of what fails.
It now floats. The pane in front carries it on a zero-height anchor pinned to
the top edge of its composer card, at the right: it never moves the
conversation, rises with the card as the prompt grows, and stays clear of
the send button. With no pane on screen it takes the main area's corner. Its
fill is mixed into the surface, since it now sits over content.
perf/errbar.mjs holds the geometry at two widths: the bar appears and takes
clicks, sits above the card, leaves send clickable, and does not move the
composer. It reaches the bar the way a real failure does, through a hub
method the bench can now refuse. Against the old in-flow bar it fails on the
composer moving 45px and on the bar sitting below the card.
main.tsx has loaded studio.css since the studio interface landed, and the
perf harness never followed: every browser guard since has measured a page
without the studio layout, with the composer at the top of the window and
the sidebar unstyled. Run against the real layout, fifteen guards still
report failures; four of them (cards, idle, press, surface) were passing
only because the layout was missing, and contrast and pick stop failing.
host_authorities is user-global: loadForRoot holds the user's value and puts
it back after the project reasonix.toml merges. The held value was the slice
header, and decoding the project file writes into the backing array the
user's decode left behind, so the held copy read back the repo's list. A
cloned repo declaring host_authorities = ["docker"] got the container daemon
socket inside the bash sandbox — with an explicit user grant and with the
default one alike.
The held value is now a copy.
RewrittenTestCriteria read Go test files only, so a Python run that moved
its own expectation reached the completion summary, the best_of_n judge and
the integrity bench with a clean record.
pytest modules are now read by pytest's default collection contract
(test_*.py / *_test.py, module-level test* functions, test* methods of
Test* classes). Bodies come from Python's own line structure — the
indentation stack, bracket joining, whole strings — and compare as token
signatures, so reformatting, comments, docstrings and formatter trailing
commas do not read as rewrites while a changed expectation or parametrize
case does. Source that does not lex claims nothing.
Detection only: holding a criterion is still Go-only, now enforced at
captureRewrittenCriteria too, because no backend re-runs a captured Python
test and holding one would owe the turn a baseline run nothing can settle.
Opening the card refreshed the share status and, in the same commit, minted
a pairing code on the status it already held. When phone access had been
turned off elsewhere since, the mint reached the kernel first and was
refused as share.closed, and the window's error bar told a person who had
only opened the card to "turn phone access on first".
Opening now waits for the read and mints only when the door is open and no
live code is held. A failure inside the card is said inside the card, under
its header, and clears when the card closes, rather than landing in the
window's error bar at the far end of the screen.
Each candidate kernel already emitted a CompletionSummary at turn end —
verdict, checks passed/failed/suppressed, review, gaps, and existing tests
it rewrote or removed — but the candidate sink forwarded usage only, so the
judge compared attempts on their final message and diff alone.
candidateSink now keeps the last summary per attempt; the judge sees it as
a host record the attempt cannot write, and the parent's report carries it
per attempt. A failed check counts against an attempt, and one that
rewrote tests the task did not ask to change is suspect however green.
Discovery searched /Applications as a literal, so the tests that stand for a
machine with no editor could not build one: on a Mac with VS Code installed
TestDiscoverRefusesWithAnIdentityWhenNothingIsInstalled found it,
TestInstalledPathsAreSearchedBeforePath found it before the placed copy, and
serve's TestEditorOpenSeparatesMissingFromFailedToStart launched it against
a controller with no workspace and got a 500. CI runners have no editors, so
all three were green there and red on a developer's machine.
Discovery now reads the machine through a host value: the OS, the
environment and the system application folder. Discover passes the real
one; the editor tests pass an empty folder. The serve test no longer tries
to empty the machine at all: an editor the user configured that does not
exist is the missing class everywhere, and it reaches the same refusal.
The sidebar's version rows had no rendering test: the expander, the label
and opening the version's own path are now exercised through the real
Workspaces component.
The lease-alignment race posted to /fork, removed with explicit forking, so
half of it only collected 404s. /new rotates the active path through another
handler, which is what the race is for.
Editing an earlier message, regenerating a reply and rewinding the
conversation all truncate the transcript, and only the latest one could be
undone. What was said before the cut was gone.
A conversation rewind now saves what it is about to cut as a version of the
session: a whole session file, marked Superseded with the session as its
parent, left out of every session list so it never reads as a second
conversation. The Studio sidebar hangs versions under their conversation in
the same expander as recovery copies, labelled apart, and each opens as a
whole conversation. A version someone carries on becomes a conversation of
its own: the first save that adds a turn clears the mark and keeps the
parent. Undoing a rewind drops the version that now duplicates the restored
conversation, judged by content so it holds after a restart. Removing a
conversation removes its versions. Keeping a version is a record, not a
precondition: a rewind that cannot keep one still happens.
NewSessionPath moves to session_path.go to keep save.go inside its budget.
Closing the manager cancelled every running job the same way kill_shell
does, so a job read after a restart came back as killed. Nobody killed it:
the process running it ended, which only the host knows, and a bare "killed"
left the model to guess between that and a stop someone asked for.
A job cancelled because its manager is closing now ends Interrupted, the
status an abandoned record already repairs to, and one stopped by Kill stays
Killed. bash_output and wait render the cause after an interrupted status.
The tool descriptions are unchanged, so the cached prefix does not move.
The boot-level restart test starts a background job, restarts, and reads
it: a restart cannot keep the job alive, but the model is told why it ended.
Status and its cause move to status.go.
The store kept pending results in memory and removed every worktree when the
session closed. A restart between delegating an isolated task and applying
it lost the work, and the refusal that followed said no result was pending,
a claim the host could not know was true.
The worktree is now the result and the store only indexes it. A finished
run writes an entry mark beside its worktree naming the id, the workspace
and the snapshot; what it changed is derived from the worktree on every
read, so no stored tree can outlive the objects it names. A lookup the
memory misses is recovered from disk, closing a session no longer removes
anything, and a result ends only when applied or discarded. Results taken
from another workspace are never offered, since applying one would write
into that workspace from this one. Unsettled results older than fourteen
days are swept at boot, and only directories carrying the mark are touched.
The boot-level restart test delegates in one turn and applies in the next,
with and without a restart between them, and both arms now land the file.
A goal resumed after a restart re-demanded inspection, verification and
sign-off for a change that had all three before the process ended: their
receipts lived in the ledger, which does not cross the boundary. It was
fail-closed but not restart-equivalent, and every restart cost a re-run.
The delivery checkpoint now carries Proven: what the turn's last readiness
verdict had proven for the pending change, with a content fingerprint of
every path the scope changed. On resume, with no new write in this process,
the proof discharges what it covers, but only while every path still holds
the content it was earned on. A path edited or deleted while no process was
watching voids all of it, which is more than an uninterrupted run can see.
A change with a part the host cannot name carries no proof at all.
The verdict is recorded where the turn asks to stop, so the checkpoint reads
the same derivation the gate used rather than a second one. The Delivery
branch of the readiness check moves to appendDeliveryGaps to keep the
function inside its complexity budget. The restart harness no longer pins
any stricter step: both arms now owe exactly the review.
The repository now describes the two lines the version roadmap announced
(discussions/10748): 2.x on studio, 1.x on main-v2 in maintenance.
- README (en, zh-CN): a Versions section and a Studio 2.x release badge.
Install leads with Reasonix Studio 2.x from the studio-v* pre-releases,
with the 1.x CLI and desktop app beside it as the stable line. The
zh-CN build-from-source section gains the Studio build it lacked.
- Issue forms offer 2.x / 1.x / 0.x, and the labeler maps them to v3 /
v2 / v1, with the pasted version still overriding the dropdown. The
forms link the roadmap. The PR template and CONTRIBUTING say which
branch takes which changes.
Issue forms, schedules and Dependabot are read from the default branch
only, so these take effect when studio becomes it:
- CodeQL scans pushes and PRs on both lines, and the weekly scan checks
out each line and files results under its own ref, so 1.x keeps its
security scanning.
- v1-maintenance-schedule starts main-v2's nightly app-memory and
diagnostic-overhead monitors, which measure the 1.x desktop app, on
main-v2.
- Dependabot drops /desktop/frontend and /site, which do not exist on
studio, and covers /desktop/frontend-next and /desktop/electron. It
also names main-v2 as a target for patch updates only.
layer-order ranked model, tools, safety, ext, state and platform as one peer
pool that could import freely inside itself, so state -> tools or
model -> platform passed the gate and nothing held the directory graph
acyclic. The pool is split into the order the tree already follows
(safety/state, model, platform, ext/tools), and a package now imports only
its own directory or a strictly lower rank. The tree needed no change.
The structured review a high-risk change owes was derived from mutation
receipts in the evidence ledger, which has no persisted form. A goal
restarted between its turns resumed with an empty ledger, derived nothing,
and completed without the review. The restart-equivalence tests showed it
through the real boot and the controller's own resume.
The delivery checkpoint now carries OwedReview: the review kinds a change in
its scope still owes and the paths a review has to cover. Each turn's
checkpoint joins what the live ledger derives with what was carried in and is
still unsettled, so a restarted ledger deriving nothing is never read as a
discharge. The gate owes a carried review until a review of that kind covering
those paths lands; one kind does not settle the other. The live derivation is
unchanged and shares one helper with the checkpoint, so an uninterrupted run
is held exactly as before.
The restart harness now fails any step where a restart owes less than an
uninterrupted run. What a restart still re-demands — inspection, verification
and sign-off, whose receipts do not cross the boundary either — is fail-closed
and pinned as known, and the boot-level test asserts both arms end blocked.
The production form of the restart-equivalence harness: two real boots, the
session resumed from its file the way --continue does, and goal turns
submitted the way a window submits them, so they carry the goal's delivery
scope. The scripted model acts on milestones rather than turn counts, and
every reply spends budget, so a goal that cannot finish stops on its own
token budget as a real one does.
A high-risk change (a path the project declares sensitive) is made, checked
and signed off, with no review. Uninterrupted, the goal stays blocked and the
host keeps demanding the review. Restarted between its two phases, the
resumed goal is never asked for one and completes. Each arm run in its own
process gives the same result, so no package state carries between them.
The review obligation is derived from mutation receipts in the evidence
ledger, which has no persisted form: the goal sidecar carries only the
DeliveryCheckpoint. The case is pinned as a known divergence so the fix turns
it red.
Past IndexMaxChars the listing was cut mid-rune: with 91 skills only 35
lines survived whole, the other 56 were invisible to the model, and the cut
could leave half an identifier to call. The listing now degrades to names
only, cut whole-line with a count of what it left out, and points at
use_capability search for descriptions. A catalog under the cap renders the
same bytes as before.
No per-turn retrieval over the user's wording: ranking by prose would hide
skills inside a gate. The model asks through the existing capability search,
whose keywords now include the triggers a skill's author declared.
Stopping the process at a turn boundary and resuming must leave the host
deciding exactly what it would have decided had it kept running. The harness
plays a scenario twice, once in one agent and once through a restart after a
chosen step, and compares the readiness verdict of every step. The restarted
arm is built as a resume builds it: a fresh agent with an empty ledger, the
transcript, and the DeliveryCheckpoint after a JSON round trip, which is all
the goal sidecar carries; the controller's own restore of that checkpoint is
established by goal_resume_project_check_test.
A signed-off ordinary change agrees across a restart. A high-risk change does
not: the uninterrupted arm owes structured_review at every later step, while
the restarted arm owes re-verification and sign-off instead, and once those
are redone it finalizes with no review at all. The review obligation is
derived from mutation receipts in the ledger, and the ledger does not cross
the boundary. The case is pinned as a known divergence, so the fix turns it
red and takes the mark off.
With [agent] worktree_isolation on, task takes isolation "worktree". The
sub-agent then runs as an unattended kernel of its own, rooted at a fresh
worktree of the workspace as it is now, and nothing it writes reaches the
workspace. Its result lists the changed paths under an iso_ id, and the
parent settles them with apply_isolated or discard_isolated.
apply_isolated declares every path it writes, so the host reserves,
checkpoints and records them like any other write, and it applies path by
path: a path the workspace changed meanwhile refuses the whole apply by code
and keeps the result. A workspace git cannot snapshot is refused with
isolation.not_git; isolation is never faked with the shared workspace.
The child runs under the session's own approval posture, read when it starts.
It has no one to ask, so a session that asks gets a child that refuses what
needs approval, and nothing it does exceeds what the session would do
unattended. Arguments a separate kernel cannot honour (profile, tools,
write_paths, continuation, background) are refused rather than dropped. The
child offers neither isolation nor best_of_n, and the settlement tools are
hidden from sub-agents. Closing the session removes every worktree it holds.
best_of_n's attempt runner is shared as runUnattended, and task's name,
description and schema move to task_schema.go.
The host observes a write through Previewer, which describes one file. A
writer that changes several paths named nowhere in its arguments was left an
opaque writer: a coverage gap on the checkpoint, no preimages, and a receipt
with no paths, so neither rewind nor the evidence ledger could say what it did.
tool.WritePathDeclarer lets such a writer name every path before it runs. The
host reserves exactly those paths against background writers, captures each
one's preimage and after-fingerprint, and records them on the receipt. A code
rewind then restores a modified path and removes a created one.
Apply refuses unless the whole workspace still equals the snapshot, which
holds for best_of_n, where the parent waits, and never for a candidate that
ran beside a parent that kept editing. ApplyPaths judges each changed path
alone: the workspace's version is hashed through the same filters a snapshot
stages with, and compared against the snapshot's blob and the candidate's.
A path still at the snapshot is written, one already at the candidate's
version is left alone, and anything else is a conflict. One conflict writes
nothing, and the ConflictError names every path.
ChangePaths refuses a change outside the folder the snapshot was taken of,
and DiffStat gives per-path line counts for reporting a candidate.
tools/list, resources/list and prompts/list took the first page and ignored
nextCursor, so a server that paginates lost every tool past its first page
without any error. The three now read every page through one helper, bounded
at 64 pages; a cursor handed out twice is reported as a server loop rather
than served as a complete catalog.
The readiness gate read only the checks the current process declared. A Goal
restored into a process whose declaration had changed carried its captured
baseline in the delivery checkpoint, but nothing enforced it, so running the
replacement alone was enough to finalize. The Goal resume path was shown
reachable by goal_resume_project_check_test.
The gate now also owes every baseline check no current declaration names,
inside the delivery scope that captured it, anchored on the same write the
gate already uses. The trust-domain rule settles the authority question: the
workspace cannot supersede a captured criterion.
The full switch to ledger obligations is not taken: a resumed process starts
from an empty ledger, where the obligations owe nothing at all, so the switch
would open the resume path this closes. A restored checkpoint from another
scope does not bind an unscoped turn.
The 2.20.0 release spent 226s packaging Windows and 640s mirroring to R2.
Packaging compressed the same 245 MB tree twice, the 7-Zip archive and
then the NSIS installer. They now run at once; the zip reads its own copy of
win-unpacked because NSIS writes elevate.exe into the tree it packs, and a
zip read beside that would contain it or not by timing. A local cross-build
gives the same 128-file zip as the sequential step.
Defender real-time scanning is switched off for the Windows build, which
writes that tree several times over. continue-on-error keeps a runner image
that refuses the switch from failing the release.
The R2 mirror sent 18,755 delta chunks of about 10 KB with the AWS CLI's
default of 10 requests in flight, roughly 31 a second. Raising it to 64 leaves
the published objects unchanged.
2026-09-24 09:06:52 -07:00
5508 changed files with 440951 additions and 35713 deletions
Target branch: 2.x features and fixes go to `studio`. 1.x takes bug fixes,
provider/API compatibility, release/updater, security and platform stability
only, on `main-v2`.
-->
## Summary
-
## Cause
-
## Blast radius
-
## Neighbouring behaviours tested
-
## Why this layer
-
## UI changes
<!--
Delete this section if the PR touches nothing under `desktop/frontend-next/src`
or the website pages.
-->
- Screenshots, before and after: light and dark, a wide window and a phone-width one (about 390 px). Attach them to this description.
- Entries moved, hidden or removed: none / list each one.
- Hiding or removing a visible control needs a linked issue and a release-notes line naming it. Do not retire UI with `display: none !important` or a feature flag.
- If a row of `docs/STUDIO_PARITY.md` changes, update it here.
## Issues
<!--
@@ -20,3 +54,7 @@ the source-to-target contract here: easing syntax, time units, callbacks,
cancellation, reduced-motion behavior, and failure fallback. Verification must
assert that the target API was actually called; a mock that silently skips it
does not count.
## Disclosure
- [ ] AI assistance was used for this change (optional; for transparency only, it does not change how the change is reviewed)
Write-Output "::error title=studio-signing.unexpected-signer::STUDIO_SIGNING_SUBJECT is '$($env:EXPECTED_SUBJECT)' but $file is signed by '$($certificate.Subject)'"
$failed = $true
}
}
if ($failed) { exit 1 }
'Certum payload and installer signing, signer, trust and timestamps passed.' >> $env:GITHUB_STEP_SUMMARY
platform=$(git diff -z --name-only --no-renames "$BASE_SHA" HEAD | bash scripts/ci-platform-scope.sh)
if git diff --name-only --no-renames "$BASE_SHA" HEAD |
grep -qE '^(desktop/electron/|desktop/packaging/|desktop/cmd/update-helper/|desktop/cmd/studio-manifest/|desktop/go\.(mod|sum)$|cmd/reasonix-studio-host/|internal/platform/(update|appupdate)/|go\.(mod|sum)$|scripts/(linux-install-smoke|verify-linux-package|verify-studio-artifacts)\.sh$|scripts/(windows-signing-lib|verify-windows-package|verify-windows-authenticode|install-smoke)\.ps1$|\.github/workflows/(studio|release-studio)\.yml$)'; then
# The update channel is the signed manifest. A second answer to which
# version is current is not a convenience. The .blockmap is the same
# channel's differential index and carries the release prefix, so it also
@@ -233,36 +592,13 @@ jobs:
echo "page: ${page:-<missing>}"
test -n "$host" -a -n "$page"
# Read off the package rather than off the config that describes it. Both
# of these went wrong the same way once: electron-builder names a package
# after package.json unless told, and fpm records the owner it finds on
# disk. Either one leaves a Studio that cannot update itself on Linux and
# says nothing about it until a user asks.
- name:The .deb is Studio's own package and carries what updates it
- name:Linux packages carry their declared architecture and update payload
if:runner.os == 'Linux'
run:|
deb=$(ls dist/ReasonixStudio-linux-*.deb)
name=$(dpkg-deb -f "$deb" Package)
if [ "$name" != "reasonix-studio" ]; then
echo "::error::the package is $name; dpkg holds that beside the installed reasonix-studio instead of upgrading it, and update.StudioLine asks for it by name"
exit 1
fi
dpkg-deb -c "$deb" > contents.txt
# The path is the polkit action's exec.path annotation, so it is the
# only one authorization is granted for; root because polkit will not
# run a helper the invoking user could have written. Either spelling
# of root: dpkg-deb prints the numeric ids it cannot map, and which
# one it prints is a property of the machine reading the package
# rather than of the package.
if ! grep -qE '^-rwxr-xr-x (root/root|0/0) +[0-9]+ .* \./usr/lib/reasonix-studio/reasonix-studio-update-helper$' contents.txt; then
echo "::error::no root-owned executable update helper at the path the polkit action names"
grep reasonix-studio-update-helper contents.txt || echo "(not in the package at all)"
exit 1
fi
if ! grep -q './usr/share/polkit-1/actions/io.reasonix.studio.update.policy$' contents.txt; then
echo "::error::the polkit action is not in the package, so pkexec has nothing to authorize against"
if ($run.ExitCode -eq 0) { throw 'the arm64 installer exited 0 on an x64 machine' }
if (@(Get-ChildItem -Recurse -Force $target -ErrorAction SilentlyContinue).Count -ne 0) { throw 'the arm64 installer wrote into the target on an x64 machine' }
<a href="https://qm.qq.com/q/i59b0z2R8s"><img src="https://img.shields.io/badge/QQ%20group-1093562660-12B7F5.svg?style=flat-square&labelColor=161b22&logo=tencentqq&logoColor=white" alt="QQ group 1093562660"/></a>
</p>
<p align="center">
@@ -39,23 +43,83 @@
<br/>
<p align="center"><strong>Open source · MIT · a single Go binary</strong></p>
<h3 align="center">A coding agent you can leave running.</h3>
<p align="center">One local engine, four ways in — terminal, desktop app, browser, or your editor over ACP. Plan mode, permissions, a workspace sandbox and per-turn checkpoints keep a long autonomous run something you can still read and undo.</p>
<p align="center"><strong>Open source · MIT · runs on your computer, model of your choice</strong></p>
<h3 align="center">An open-source coding agent for your terminal, desktop, browser and editor.</h3>
<p align="center">It reads your project, edits files and runs commands and tests. Have it ask before each step, or work on its own within limits you set. Files its edit tools changed can be rewound. New here? Start with the Studio desktop app.</p>
<p align="center">Maintained by <strong>Huahui Yu</strong>.</p>
> [!IMPORTANT]
> **Community · 加入社区** — bilingual Discord for setup help (`#help` / `#求助`), workflow showcases, and feature ideas. → **<https://discord.gg/XF78rEME2D>**
>
> QQ group **DeepSeek-Reasonix官方群** (`1093562660`) → **<https://qm.qq.com/q/i59b0z2R8s>** ([QR code](./docs/assets/qq-group.svg))
> Douyin: **做游戏的小鱼** (Douyin ID `22703872788`) · [QR code](./docs/assets/douyin.png)
<br/>
## ❤️Sponsor
<table><tr><td width="180" align="center" valign="middle"><a href="https://www.rightapi.ai/register?aff=e083bf73" target="_blank" rel="sponsored noopener noreferrer"><img src="./.github/sponsor/rightcode.png" alt="Right Code" width="64"></a></td><td valign="middle">Thanks to Right Code for sponsoring this project! Right Code reliably provides routing services for models such as Claude Code, Codex, and Gemini, with both pay-as-you-go and monthly subscription billing options available. Invoices are available upon top-up, and enterprise and team users can receive dedicated one-on-one support. Register via <a href="https://www.rightapi.ai/register?aff=e083bf73" target="_blank" rel="sponsored noopener noreferrer">this link</a>.</td></tr><tr><td width="180" align="center" valign="middle"><a href="https://www.packyapi.ai/register?aff=rIbQ" target="_blank" rel="sponsored noopener noreferrer"><picture><source media="(prefers-color-scheme: dark)" srcset="./.github/sponsor/packycode-dark.svg"><img src="./.github/sponsor/packycode-light.svg" alt="PackyCode" width="80"></picture></a></td><td valign="middle">Thanks to PackyCode for sponsoring this project! Access leading AI models through PackyCode with one API endpoint and one API key, with automatic failover and dedicated high-speed routes for Codex and Claude Code. Register via <a href="https://www.packyapi.ai/register?aff=rIbQ" target="_blank" rel="sponsored noopener noreferrer">this link</a>.</td></tr></table>
## What it is
Reasonix is an open-source AI coding assistant. The program runs on your computer; the AI model is a service you configure, which receives what it reads. You describe a task in plain language; it reads code, edits files and runs commands and tests inside the project folder you choose, and shows you each step.
You pick the model: a DeepSeek preset is built in, any OpenAI-compatible service is one config entry, and you pay the model provider directly.
## What you get
- **Read, edit, run.** It reads code, edits files and runs tests and commands, all inside your project folder.
- **Permission levels.** Have it ask every time, or let everyday actions run on their own; your forbid rules always hold. To see the approach first, use plan mode: it starts work only once you approve.
- **Rewind.** A file is saved before it changes, so you can rewind by conversation turn without touching your git history. Only files its edit tools changed are covered, not what a shell command did.
## 30-second start
Download, add a model key, open a project. New here? Start with the Studio desktop app.
1. Download and install the desktop app, Reasonix Studio, from the [releases page](https://github.com/esengine/DeepSeek-Reasonix/releases?q=studio-v&expanded=true).
2. Open it and enter your model provider and key when asked.
3. Open your project folder and write what you want, for example "add retry on failure to the http client".
4. At the level you chose, it stops and asks before anything that needs your say-so. Allow it to continue, or refuse.
Prefer a terminal? Download the `reasonix` command-line tool, run `reasonix setup` to pick a model, then run `reasonix`. See [Install](#install) and [Quick start](#quick-start).
## Four ways to use it
- **Desktop app (Studio).** A graphical app; the best place to start.
- **Terminal.** A full-screen text interface; `reasonix run "task"` also works in scripts.
- **Browser.** `reasonix web` opens a web interface on your own machine.
- **Editor.** Install the extension in VS Code; other editors that support ACP (Agent Client Protocol, a common way for an editor to connect to an AI assistant) can connect to `reasonix acp`.
All four are the same program underneath. What follows is for advanced users and contributors.
## Versions
Reasonix ships on two lines. See the [version roadmap announcement](https://github.com/esengine/DeepSeek-Reasonix/discussions/10748) for the reasoning behind the split.
| Line | Branch | Status | Get it |
| --- | --- | --- | --- |
| **Reasonix 2.x** | `studio` (this branch) | Active development | [Studio releases](https://github.com/esengine/DeepSeek-Reasonix/releases?q=studio-v&expanded=true) |
- HTTP `serve` protections for the unauthenticated local server, including
localhost binding assumptions, JSON-only state-changing requests, and CORS
restrictions.
- HTTP `serve` protections, including the default per-launch token, the launch
token every state-changing request requires when authentication is opted
out, localhost binding assumptions, JSON-only state-changing requests, and
CORS restrictions.
- Desktop and bot session isolation, including per-workspace session metadata
and configured bot allowlists.
- Updater, install, and release verification paths.
@@ -67,10 +68,10 @@ The following are normally treated as trusted local/operator-controlled inputs
unless another bug lets an untrusted actor supply them:
- CLI arguments and text typed directly by the local user.
-Project configuration files intentionally loaded from the current workspace.
-The user's own configuration under the Reasonix home directory.
- Explicit `@path` references supplied by the local user to attach local files.
- MCP servers, language servers, hooks, and slash commands installed or enabled
by the local user.
- MCP servers, language servers, hooks, and slash commands installed, enabled,
or approved by the local user.
- Provider base URLs and model names configured by the local user.
The following can be security issues when reachable by an untrusted actor or
@@ -89,6 +90,56 @@ when they bypass the intended boundary:
- Trusting unverified update artifacts, plugin definitions, or downloaded
binaries.
## Project Configuration
Files that arrive with a workspace are untrusted input: `reasonix.toml`,
`.reasonix/settings.json`, `.mcp.json` and the project `.env`. Cloning a
repository is not a decision to trust its author, so these files may narrow
what the user configured and never widen it.
| Setting | What a workspace file may do |
| --- | --- |
| `sandbox.bash` | Move `off` to `enforce`, never back |
| `sandbox.network` | Move `true` to `false`, never back |
| `sandbox.forbid_read`, `sandbox.denied_domains` | Add to the user's lists |
| `sandbox.allow_write`, `sandbox.workspace_root` | Apply only when they resolve inside the workspace, symlinks followed; `${VAR}` in any sandbox path expands from the process environment only |
| `sandbox.host_authorities`, a user's `sandbox.allowed_domains` | Nothing |
| `permissions.deny`, `permissions.ask` | Add to the user's rules |
| `[[providers]]`, `tools.system_one` endpoints | Declare them; they stay out of use until approved, and model settings naming them fall back to the user's |
An "always allow" answer, or extra write access, granted in a workspace is
stored under the Reasonix home, not in the workspace.
Programs a workspace names for the host to run stay off until the user
approves them for that workspace:
- Hooks, language server commands, `tools.search.rg_path`,
`tools.shell.path`, `[browser]` and `tools.system_one.laya.python` are
approved with `reasonix trust`, or for hooks by saving them from the hooks
editor.
- An approval is stored under the Reasonix home with a digest of the
declaration and of the workspace files it names; any change needs approval
again. Approvals and grants belong to one workspace folder.
- Hooks and language servers are checked against their approval again before
each start, and refused if a file they name changed, appeared or now
resolves elsewhere.
- A single program path (`tools.shell.path`, `tools.search.rg_path`,
`[browser]`, `laya.python`) may not point into the workspace, an
`allow_write` root, a temporary directory or a toolchain cache the bash jail
leaves writable, and may not climb with `..`; its approval covers the
program file's content.
-`tools.system_one.laya.local` is read from the user's configuration only.
- Without an interactive frontend they stay off.
- MCP servers a workspace declares stay off until the user enables them for
that workspace.
A project value that is not applied is reported as a load warning with the
reason. A way for workspace files to widen any of the above without the
user's approval is a security issue.
## `@` File References
Reasonix supports `@path` references so users can include local files and images
| `-task` | *(all)* | Suite mode: run only these comma-separated task IDs (e.g. `-task fix-add-bug`); unknown IDs fail with the available list. |
| `-attempts` | `1` | Suite and diff modes: retry a task until an attempt passes, up to N; enables the `Pass@≤N` KPI, and TTCS charges a retried solve with its failed attempts' wall. |
| `-trials` | `1` | Suite mode: run every task N times and keep every outcome. Adds a **Reliability** line: `pass^N` (tasks that passed every trial),`pass@N` (any trial), mean `pass@1` over all trials, and the flaky tasks by name. `Pass@≤N` asks whether the agent can solve a task; `pass^N` asks whether it does so every time. Cannot be combined with `-attempts`. |
| `-trials` | `1` | Suite mode: run every task N times and keep every outcome. Adds a **Reliability** line: `pass^N` (complete tasks that passed every trial) and`pass@N` (complete tasks with any pass), both over the tasks that ran all N trials only (`n/a` when none did), mean `pass@1` over every graded trial, and the flaky tasks by name; a task the token budget cut short is listed on an **Incomplete** line and is in neither. `Pass@≤N` asks whether the agent can solve a task; `pass^N` asks whether it does so every time. Cannot be combined with `-attempts`. |
| `-tape-record DIR` | — | Suite mode, with `-meter`: record every provider exchange at the meter, one tape per task and trial (`DIR/<task>/trial-N/0001.request.json`, `.response.json`, `.response.body`). Taped runs use a fixed workdir path and pin the seeded files' times, because both reach the model through the workspace block and file listings. |
| `-tape-replay DIR` | — | Suite mode, with `-meter`: answer each run from its tape instead of the provider, and add a **Replay** line naming the first request whose body differs from the recorded one, down to the message (`request 2, message 4 (tool)`), the tool list, or a field. The diverging request is kept as `NNNN.replayed.json` beside the recorded one. A kernel change that should not alter what the model is asked replays with every request matched. |
| `-bin` | `reasonix` | Path to the reasonix binary. |
| `-profile` | `baseline` | Tool-surface/runtime tier: `baseline` \| `economy` \| `balanced` \| `delivery`. All but `baseline` append `--profile <tier>` to the agent invocation; `baseline` passes no flag (byte-identical legacy control, behaviorally `balanced`). Economy starts with the core tool set and pays `connect_tool_source` rounds plus prefix resets to grow it — the report's Tool surface line prices that trade. |
| `-json` | *(none)* | Write the JSON report here (optional). |
| `-trajectories` | *(none)* | Suite mode: write one `<task-id>.trajectory.jsonl` per task into this directory (the agent's full event stream with timestamps — see `reasonix run --trajectory`). The report gains a time-attribution line (tools vs. model) and each JSON result a `trajectory` digest. |
| `-force-planner` | `false` | Suite mode: prefix each prompt with a plan-first directive so the two-model turn engages regardless of the planner gate. Use for the "with planner" arm of an A/B; results carry `plan_forced` so arms are only comparable with equal forcing. |
| `-anchor` | `blind` | Suite mode: which hypothesis the agent holds before it looks at anything — `blind` (none, the control) \| `correct` \| `wrong`. The seeded arms prefix each prompt with the task's authored seed and **skip** tasks that have none, so an unseeded control run never lands in a seeded denominator. Results carry `anchor`. See [Anchor resistance](#anchor-resistance). |
| `-cache` | `cold` | Suite mode: `cold` runs each task as a fresh session (the fair cross-agent comparison arm); `warm` primes the provider prefix cache with a one-step run in the same workdir first, measuring the long-lived-session steady state. Never mix arms in one report — compare them with `-mode compare cold.json warm.json`. |
| `-policy` | *(none)* | Suite mode: experiment arm. `memory-off` hides the memory store, which is the MemoryBench counterfactual arm; the empty default runs the baseline. |
| `-budget` | `800000` | Abort once total tokens cross this (`0` = no cap). Remaining tasks are reported as skipped. |
| `-meter` | *(off)* | Suite mode: route the benchmarked provider through the neutral measuring proxy, using this `config.toml` as the source. Spend is then counted at the request boundary instead of trusted from the harness. See [Neutral metering](#neutral-metering). |
| `-faults` | *(none)* | Suite mode: inject provider failures through the meter — absolute indices (`3:429`) and/or a cadence that scales with the run (`every:5:500`). Requires `-meter`. See [Fault recovery](#fault-recovery). |
@@ -541,33 +542,32 @@ writer.
## context-maintenance-e2e
This harness measures what happens when a long session goes idle past the
provider's cache TTL and then resumes: it A/B-compares cold-restart miss tokens
with and without pruning, and checks that the agent re-reads a file behind a
prune placeholder instead of hallucinating.
A cost-capped smoke for content-driven maintenance, in three legs over one state
directory: seed a session just below the compaction trigger, load it back and
check nothing is re-derived, then grow it past the trigger and check exactly one
summary lands.
It is hardcoded to the `deepseek-v4-flash` model at `https://api.deepseek.com`
and requires the `DEEPSEEK_API_KEY` environment variable.
Seed and resume take no provider; `continue` samples and needs
`DEEPSEEK_API_KEY`. The model is hardcoded to `deepseek-v4-flash` at
`https://api.deepseek.com`.
```sh
export DEEPSEEK_API_KEY=...
# Seed below the trigger: no checkpoint installed, no summary started
go run ./benchmarks/context-maintenance-e2e -offline seed
# Seed both arms (pruned + control) with a large session and warm the cache
go run ./benchmarks/context-maintenance-e2e seed
# Wait past the provider's cache TTL, then resume: prune the "pruned" arm and
# compare cold-restart miss tokens
# Reload: the projection version must not move and no summary may run
go run ./benchmarks/context-maintenance-e2e resume
# Run the comprehension trials (agent must re-read a pruned file and answer
# from it); exits non-zero unless every trial passes
go run ./benchmarks/context-maintenance-e2e comprehension
# Grow past the trigger: one summary, version +1, then no further summary
export DEEPSEEK_API_KEY=...
go run ./benchmarks/context-maintenance-e2e continue
```
| Flag | Default | Purpose |
| --- | --- | --- |
| `-dir` | `benchmarks/context-maintenance-e2e/run` | State directory for `seed`/`resume` (sessions + `meta.json`, `resume-<ts>.json`). |
| `-trials` | `5` | Number of comprehension trials. |
| `-dir` | `benchmarks/context-maintenance-e2e/run` | State directory shared by the three legs (session + `meta.json`). |
| `-max-usd` | `0.50` | Hard cost cap for the live legs. |
| `-offline` | off | Skip the live provider; `seed` only. |
1. Measured on 2026-10-17 against the official `https://api.deepseek.com` OpenAI adapter, using the model returned by `/models`: `deepseek-flash`. Thinking was disabled. Inputs were synthetic; no workspace history or credentials are recorded here.
2. Each of three text and three rich-history trials sent a main request to warm the prefix, then called the actual agent summary path, then an independent summary control. Rich history included a synthetic closed `read_file` call, result and reasoning. Warmup outputs were discarded, never executed.
3. A summary had a 1,024-token output cap; warmups had 64. Success means a heading, preservation of the fixture's `ParseDuration` and zero-value constraint, no emitted tool calls, and `finish_reason=stop`. This checks the fixture's contract, not arbitrary-summary factual completeness.
4. First-attempt reuse success counts successful reused summaries / six reused summaries. Fallback rate counts independent requests initiated by the reuse coordinator / six reuse trials; the six explicit independent controls are excluded. Both were 6/6 success and 0/6 fallbacks. Independent controls also succeeded 6/6.
5. The raw result file records provider-reported usage and full stream latency, measured from adapter `Stream` until channel close. Token rows below sum only the three indicated summary requests; they exclude warmup, model listing and any earlier exploratory requests. Cache residency was uncontrolled; prior probes had warmed these synthetic prefixes.
## Results
| History | Shape | N | Prompt | Hit | Miss | Output | Median ms | Estimated USD |
1. Estimates use the [official pricing](https://api-docs.deepseek.com/quick_start/pricing/) checked on the measurement date: off-peak Flash USD per million tokens, hit 0.003, miss 0.15, output 0.6. October 17 was a Saturday. Formula: `(hit*0.003 + miss*0.15 + output*0.6)/1_000_000`. These are token estimates, not account deductions.
2. Reuse hit fractions were 82.7% (text) and 81.3% (rich), versus 63.1% and 57.3% for controls. This small fixture does **not** establish cost savings: retaining the system/catalog increased both total input and input cost. Text total cost increased 6.8%; rich total decreased 12.5% because output was shorter, not because input was cheaper.
3. The six warmups consumed 7,032 prompt and 231 output tokens, estimated USD 0.000346680. All 18 recorded generation calls together cost an estimated USD 0.002491470. No failed-digest retry occurred in these live trials; the deterministic fallback tests establish its behavior and accounting.
4. A rejected nonempty reused answer earns one independent-shape fallback. A separate truncation retry may enlarge the output cap, for at most three logical model calls total. Every actual attempt emits and records its own spend; returned usage is their sum. Empty output, stream errors and cancellation do not earn shape retries.
5. These sequential, six-pair trials establish interoperability for one model, mode and short fixture. They do not estimate production failure probability, thinking-mode behavior, other models, cold-cache cost, or large-history savings.
## Reproduction
1. Provide a test key through an existing secret environment configuration. Do not put credentials in source, command arguments or captured logs. Select a model currently returned by `/models`.
2. Run the opt-in probe; it skips without `DEEPSEEK_API_KEY`. The dedicated build tag avoids enabling unrelated live tests.
```sh
REASONIX_LIVE_MODEL=deepseek-flash go test -tags live_compaction ./internal/runtime/agent -run '^TestLiveCompactionPrefix$' -v -count=1
```
3. This probe exercises actual summary construction and the HTTP adapter. The hermetic `TestEffectCompactionSharesThePreviouslySentPrefix` separately exercises manual, automatic pressure and rich history through `boot.Build`, comparing actual JSON request-prefix and schema bytes. Its sparse-history case also proves an interleaved retained user turn survives the fold.
There are no measured results yet. Component tests and scripted providers prove
delivery, validation and replay; they do not measure a model's choice of interface
or a person's effort.
## Paired runs
Run each task twice in fresh desktop sessions: ordinary text/ask interaction and
native surfaces. Use the same model, sampling settings, initial workspace,
permission mode and task prompt. Record commit, model, configuration and the
native-surface capability setting. Alternate condition order across repetitions.
Keep the hidden tool out of the provider prefix in both conditions; only its
availability through capability discovery differs.
The baseline may batch questions or accept a single text reply. Never force it to
ask one question per field. Do not tell either model how many turns to use or
which component to render. If the surface-enabled model chooses text, retain
that run in the surface condition. Interface choice is part of the experiment.
The operator knows the private values below but does not volunteer them before
the model requests the corresponding information. Supply all requested values
together, regardless of presentation. Allow inspection and local editing before
submission. Save the trajectory and an observer ledger locally; raw transcripts
and draft content must not enter the aggregate report.
## Fixed tasks
| Task | Initial prompt, identical in both conditions | Private operator input | Completion criterion |
| --- | --- | --- | --- |
| `release-brief` | Prepare a release brief. Ask me for the audience, release date and whether this is a staged rollout before writing it. | audience: existing desktop users; date: 2026-11-03; staged rollout: yes | Final brief contains all three values and adds no release commitment. |
| `storage-choice` | Help me choose storage for a small desktop app. Present reasonable choices, ask me to choose and ask whether offline operation is required. Then summarize the decision. | storage: SQLite; offline: yes | Final decision states SQLite and offline operation, with no installation or file changes. |
| `draft-revision` | Draft a two-line release announcement and let me revise it before returning the final version verbatim. | Replace with the exact string ` Desktop release ready.\nKeep existing settings.\n` (decode the escapes). | Final returned string has exactly the specified leading spaces and newlines. Formatting wrappers are excluded from the comparison. |
| `preview-decision` | Show a plain-text before/after preview changing retry count from 2 to 3. Ask whether to keep that change and whether to include a release note. Return a decision summary; do not change files. | keep change: yes; release note: no | Preview shows 2 to 3; final summary preserves both decisions; workspace is unchanged. |
Freeze this task set before collecting results. Record a new protocol version if
prompts, private inputs, completion criteria or metric definitions change.
Neither the model's final claim nor a successful input acknowledgment establishes
completion. The observer checks the resulting output against the criterion.
## Metric definitions (version 1)
The anchor is acknowledgment of the initial task prompt. The observation ends
when the criterion is verified, the operator abandons the task, or the run fails.
The initial prompt is excluded from subsequent-input counts because both
conditions start from the same request.
| Metric | Count and population | Exclusions |
| --- | --- | --- |
| `acknowledged_inputs` | Distinct operator submissions acknowledged after the anchor: chat replies, ask answers, surface answers and surface actions. Includes acknowledged input later withdrawn or superseded. | Initial task prompt, local draft edits, retries with the same acknowledged request identity, automatic dispatch, replay and observer grading. This counts input exchanges, not provider rounds. |
| `submission_attempts` | Explicit operator send attempts, including rejected submissions and manual retries. | Automatic transport retries, replay, local edits and model tool calls. |
| `withdrawals` | Successful operator withdrawals of admitted input. | Failed withdrawals, automatic cancellation and replay. |
| `completed` | Runs whose output meets the fixed criterion. Denominator: every started run in each condition. | None; failures and abandoned runs remain in the denominator. |
Record these in a content-free ledger: task id, repetition, condition, observer
event ordinal, event type, opaque request identity, acknowledgment status and
completion verdict. Record the source trajectory reference and observed start
and end anchors alongside it. Distinguish manual observation from host-recorded
events. Do not copy labels, prompts, private values, drafts or answer strings into
the ledger. Identities are local correlation handles, not account identifiers.
The current pane counters and rendered trajectory export cannot reconstruct all
acknowledgments, refusals and withdrawals. Do not infer them from `ask` calls,
tool-result counts, final prose, message wording or card mounts. Until structural
acknowledgment records cover every channel, an observer ledger is required.
## Report and decision
Report the number started, completed, failed and abandoned for each condition,
then paired input-count differences for repetitions where **both** conditions
meet the same criterion. List excluded pairs and their reason beside that
comparison. Also report submission attempts and withdrawals; fewer admitted
inputs with more failures is not automatically less effort.
Keep per-task results visible. Do not combine scripted fixtures with observed
model/operator runs or describe this small task set as an account-wide result.
Report model/configuration differences and missing evidence as limitations.
The RFC's benefit remains unproven until observed trajectories show fewer input
exchanges to the same verified outcome. These results alone do not justify P3;
adaptive layout or MCP Apps needs a measured problem in addition to this result.
The owner-requested map, comparison and dashboard workspaces extend the delivery
scope independently of this study. Their scripted acceptance fixtures verify
rendering and interaction only; they do not count as observed benefit evidence.
"""Convert `amount` at `rate` and return the product.
Both units belong to the caller. A negative amount converts unchanged.
"""
returnamount*rate
defspread(bid,ask):
"""Return `ask - bid`, the spread between the two quoted prices.
A negative result means the bid is above the ask.
"""
returnask-bid
Loaded 100 of 5508 files, more files were not shown because too many files have changed in this diff.
Show more
Reference in new issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
esengine
