docs(readme): clarify /loop runs a prompt or custom command, not built-ins

- stopLoop re-arms the poll ticker after removing one loop, so stopping one of
  several no longer strands the rest with no tick scheduled (jatmn).
- a bare self-paced /command is rejected with a pointer to the interval form:
  self-paced drives a free-form goal via the playbook + LOOP: control line, which
  a custom command's expanded template never carries, so it would silently
  degrade to adaptive polling (jatmn).
- /resume only tears loops down on an actual session-id change; /resume latest or
  /resume <currentID> can resolve to the already-active session, whose loops are
  its own, not a "previous" session's (CodeRabbit).
- tests: skipped-behind-modal loop stays scheduled, footer carries the next-wake
  time, the confirmed /clear actually wipes the transcript, plus multi-loop stop
  keeping the ticker alive and self-paced-/command rejection.

.gitattributes

@@ -1,3 +0,0 @@
-*.go text eol=lf
-*.mod text eol=lf
-*.sum text eol=lf

.github/ISSUE_TEMPLATE/config.yml

@@ -1,11 +1,5 @@
 blank_issues_enabled: false
 contact_links:
-  - name: Questions and setup help
-    url: https://github.com/Gitlawb/zero/discussions/categories/q-a
-    about: Ask in Discussions Q&A. Include zero --version, your OS, provider/model, and zero doctor output for setup problems.
-  - name: Feature ideas
-    url: https://github.com/Gitlawb/zero/discussions/categories/ideas
-    about: Propose and discuss ideas in Discussions before opening a PR (Zero is in controlled development, see CONTRIBUTING.md).
   - name: Security vulnerability
     url: https://github.com/Gitlawb/zero/security/advisories/new
     about: Report security issues privately via GitHub Security Advisories (see SECURITY.md). Do not file a public issue.

.github/ISSUE_TEMPLATE/feature_request.yml

@@ -1,22 +1,12 @@
-name: Feature request (previously discussed)
-description: File only after your idea gained traction in Discussions Ideas or a maintainer asked for a tracking issue
+name: Feature request
+description: Propose a new capability or change
 labels: ["enhancement"]
 body:
   - type: markdown
     attributes:
       value: |
-        New feature ideas start in [Discussions Ideas](https://github.com/Gitlawb/zero/discussions/categories/ideas),
-        not here. This form is for the step after that: an idea that gained traction in Discussions,
-        or one a maintainer asked to have tracked as an issue. Please link the discussion below.
-        Do not begin implementation until the issue has been reviewed and marked with the
+        Please do not begin implementation until the issue has been reviewed and marked with the
         `issue-approved` label (see CONTRIBUTING.md).
-  - type: input
-    id: discussion
-    attributes:
-      label: Ideas discussion
-      description: Link to the Discussions Ideas thread (or note which maintainer asked for this issue).
-    validations:
-      required: true
   - type: textarea
     id: problem
     attributes:

.github/workflows/ci.yml

@@ -30,36 +30,6 @@ jobs:
           go-version-file: go.mod
           cache: true
 
-      - name: Install Make on Windows
-        if: matrix.os == 'windows-latest'
-        shell: pwsh
-        run: choco install make --no-progress -y
-
-      - name: Validate quality targets with stock macOS Make
-        if: matrix.os == 'macos-latest'
-        env:
-          GIT_CONFIG_COUNT: 1
-          GIT_CONFIG_KEY_0: grep.patternType
-          GIT_CONFIG_VALUE_0: fixed
-        run: make -n vulncheck
-
-      - name: Check vulnerabilities through native Windows Make
-        if: matrix.os == 'windows-latest'
-        shell: cmd
-        run: make vulncheck
-
-      - name: Check dead code through native Windows Make
-        if: matrix.os == 'windows-latest'
-        continue-on-error: true
-        shell: cmd
-        run: make deadcode
-
-      - name: Run static lint through native Windows Make
-        if: matrix.os == 'windows-latest'
-        continue-on-error: true
-        shell: cmd
-        run: make lint-static
-
       - name: Check formatting
         if: matrix.os == 'ubuntu-latest'
         shell: bash
@@ -130,18 +100,27 @@ jobs:
           go-version-file: go.mod
           cache: true
 
+      # govulncheck, deadcode, and golangci-lint each resolve a toolchain from
+      # their own modules, which can drop below the version go.mod requires and
+      # then fail to load our packages. Pin GOTOOLCHAIN to the go.mod toolchain
+      # so all three run under it.
+      - name: Pin toolchain from go.mod
+        run: |
+          toolchain="$(awk '/^toolchain /{print $2}' go.mod)"
+          echo "GOTOOLCHAIN=${toolchain:-auto}" >> "$GITHUB_ENV"
+
       # Hard gate: fails the build when code reaches a known vulnerability. A stdlib
       # CVE is cleared by a toolchain bump (see go.mod). May also flag a newly
       # published advisory on an unrelated PR — intentional: do not ship known vulns.
       - name: govulncheck
-        run: make vulncheck
+        run: go run golang.org/x/vuln/cmd/[email protected] ./...
 
       # Advisory: reports functions unreachable from any cmd/* main so dormant
       # code is visible in CI. Non-blocking while the dormant subsystems are
       # still being wired or removed.
       - name: deadcode (advisory)
         continue-on-error: true
-        run: make deadcode
+        run: go run golang.org/x/tools/cmd/[email protected] -test=false ./...
 
       # Advisory: catches what deadcode's whole-program reachability analysis
       # doesn't, unused private functions/assignments reachable within a
@@ -151,4 +130,4 @@ jobs:
       # the repo are cleaned up incrementally (see #527).
       - name: golangci-lint (advisory)
         continue-on-error: true
-        run: make lint-static
+        run: go run github.com/golangci/golangci-lint/v2/cmd/[email protected] run --enable-only unused,ineffassign,staticcheck ./...

.github/workflows/release-artifacts.yml

@@ -97,21 +97,16 @@ jobs:
           gh release create "$tag" --repo "$GITHUB_REPOSITORY" --title "$tag" --generate-notes 2>/dev/null || true
           gh release upload "$tag" dist/release/* --repo "$GITHUB_REPOSITORY" --clobber
 
-  # Publishes @gitlawb/zero to npm: first the five platform payloads (versions
-  # of the SAME package at suffixed versions like X.Y.Z-linux-x64, carrying the
-  # native binary + vendored helpers), then the wrapper at X.Y.Z whose
-  # optionalDependencies alias those platform versions. Platform versions MUST
-  # publish first — the wrapper pins them exactly, so the reverse order would
-  # open a window where installs resolve aliases that 404. See
-  # docs/NPM_PACKAGING.md for the full contract.
+  # Publishes the npm wrapper (@gitlawb/zero) once every platform's release
+  # assets are live — the package's postinstall downloads its binary from the
+  # GitHub Release, so publishing before the assets exist would ship a package
+  # that fails to install.
   #
   # Auth is npm OIDC trusted publishing: no NPM_TOKEN secret. The package's
   # npmjs.com settings must list this repo + workflow file as a trusted
   # publisher, and provenance is then generated automatically. Note npm only
   # allows configuring a trusted publisher on an EXISTING package, so the very
   # first @gitlawb/zero publish is a one-time manual `npm publish` bootstrap.
-  # Because the platform payloads are versions of the same package, they reuse
-  # that single trusted-publisher configuration.
   publish-npm:
     name: Publish npm package
     needs: package
@@ -161,11 +156,10 @@ jobs:
             console.log(`npm ${version} supports trusted publishing.`);
           '
 
-      # Probe the exact URLs the wrapper's first-run downloader builds
-      # (internal/release/release.go asset-name scheme) as an anonymous client
-      # — install.sh/install.ps1 and `zero update` read them too. This also
-      # blocks publishing while the repo is still private: private release
-      # assets 404 without auth, exactly as they would for an installing user.
+      # Probe the exact URLs postinstall will build (internal/release/release.go
+      # asset-name scheme) as an anonymous client. This also blocks publishing
+      # while the repo is still private — private release assets 404 without
+      # auth, exactly as they would for an installing user.
       - name: Verify release assets are downloadable
         shell: bash
         run: |
@@ -179,64 +173,12 @@ jobs:
             "zero-v${version}-windows-x64.zip"; do
             for url in "${base}/${asset}" "${base}/${asset}.sha256"; do
               if ! curl -fsSLI -o /dev/null "$url"; then
-                echo "::error::${url} is not publicly downloadable — the wrapper's fallback downloader and install.sh would fail. Is the repo public and are all assets uploaded?" >&2
+                echo "::error::${url} is not publicly downloadable — npm postinstall would fail. Is the repo public and are all assets uploaded?" >&2
                 exit 1
               fi
             done
           done
           echo "all release assets for v${version} are publicly downloadable."
 
-      - name: Download release artifacts
-        uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v4.3.0
-        with:
-          pattern: zero-*
-          path: dist/artifacts
-          merge-multiple: true
-
-      - name: Assemble npm packages
-        run: node scripts/npm/build-platform-packages.mjs --artifacts-dir dist/artifacts --out-dir dist/npm
-
-      # Platform versions are semver PRERELEASES of the wrapper version
-      # (X.Y.Z-linux-x64), so they must never publish under the default
-      # `latest` dist-tag — that would make `npm install @gitlawb/zero`
-      # deliver a platform payload as the CLI. Publish them under a dedicated
-      # tag and assert `latest` survived before the wrapper goes out.
-      - name: Publish platform packages to npm
-        shell: bash
-        run: |
-          version="$(node -p "require('./package.json').version")"
-          for dir in dist/npm/platforms/*/; do
-            echo "publishing ${dir}"
-            (cd "$dir" && npm publish --access public --tag platform)
-          done
-          # Tolerate a failed lookup (registry lag, or no latest tag yet on a
-          # bootstrap publish): under set -e a bare failure here would abort
-          # AFTER the platform payloads went out, leaving no wrapper and a
-          # re-run that fails on republishing existing versions. An empty
-          # value falls through the guard below.
-          latest="$(npm view @gitlawb/zero dist-tags.latest || true)"
-          case "$latest" in
-            "${version}-"*)
-              echo "::error::dist-tag latest points at platform version ${latest} — a platform publish clobbered it. Fix the dist-tag before the wrapper publish." >&2
-              exit 1
-              ;;
-          esac
-          echo "dist-tag latest is ${latest} (unchanged by platform publishes)."
-
-      - name: Publish wrapper to npm
-        shell: bash
-        run: |
-          version="$(node -p "require('./package.json').version")"
-          (cd dist/npm/wrapper && npm publish --access public)
-          # Confirm the wrapper owns `latest` (retry: registry metadata can lag
-          # a publish by a few seconds).
-          for attempt in 1 2 3 4 5; do
-            latest="$(npm view @gitlawb/zero dist-tags.latest || true)"
-            if [ "$latest" = "$version" ]; then
-              echo "dist-tag latest is ${version}."
-              exit 0
-            fi
-            sleep 5
-          done
-          echo "::error::dist-tag latest is ${latest}, expected ${version} after the wrapper publish." >&2
-          exit 1
+      - name: Publish to npm
+        run: npm publish --access public

.gitignore

@@ -27,10 +27,6 @@ report.[0-9]_.[0-9]_.[0-9]_.[0-9]_.json
 .cache
 *.tsbuildinfo
 
-# Go build/module caches (GOCACHE/GOMODCACHE pointed here for local runs)
-.gocache/
-.gomodcache/
-
 # IntelliJ based IDEs
 .idea
 
@@ -41,10 +37,7 @@ report.[0-9]_.[0-9]_.[0-9]_.[0-9]_.json
 
 /zero
 /zero.exe
-/zero-windows-command-runner.exe
-/zero-windows-sandbox-setup.exe
 /zero-skills
-/.zero-binary-version
 /snake-game.html
 
 # Superpowers skill artifacts (brainstorm mockups, specs, plans) — local only
@@ -54,7 +47,3 @@ docs/superpowers/
 # Termux/Android build artifacts
 zero-termux-arm64
 zero-linux-sandbox
-zero-seccomp
-
-# Generated benchmark reports are evidence for one configuration, never tree state
-internal/perfbench/reports/*.json

.release-please-manifest.json

@@ -1 +1 @@
-{".":"0.7.0"}
+{".":"0.1.0"}

AGENTS.md

@@ -1,105 +1,326 @@
-# Repository Guidelines for Zero
+---
+description: Extending Zero — how to write AGENTS.md files, custom specialists, skills, hooks, MCP servers, and plugins for the open-source CLI coding agent.
+globs: "*.go, *.js, *.md, *.json, *.toml, *.yaml, *.yml"
+alwaysApply: false
+---
 
-These instructions apply to all work in this repository. For the user-facing
-guide to extending Zero with specialists, hooks, plugins, MCP, and skills, see
-[docs/EXTENDING.md](docs/EXTENDING.md).
+# Extending Zero
 
-## 1. Contribution and Pull Request Rules
+Zero is an open-source terminal coding agent. Out of the box it does the obvious things — read, edit, run, search — but the design point of the project is that **every surface is configurable**. This document is the user-facing guide for that configuration.
 
-- Before opening any pull request, **all contributors**—including maintainers,
-  community contributors, and coding agents—must read and follow
-  [CONTRIBUTING.md](CONTRIBUTING.md).
-- Community pull requests require an existing parent issue with the
-  `issue-approved` label. Team members may open pull requests through the
-  internal development process described in `CONTRIBUTING.md`.
-- Keep each change focused on the approved or assigned scope. Do not include
-  unrelated fixes, refactors, formatting churn, generated output, or existing
-  local changes in the same commit or pull request.
-- Discuss new implementation languages, runtimes, major dependency changes,
-  and broad architectural rewrites with maintainers before implementation.
-- Pull request descriptions must explain what changed and why, link the parent
-  issue when required, and list the tests or verification performed. Include
-  screenshots or a short recording for user-visible UI changes when practical.
+If you only want to *use* Zero, the [README](README.md) is enough. This page is for the other three jobs:
 
-## 2. Repository and Implementation Conventions
+1. Tell the agent about *your* project (drop an `AGENTS.md` in your repo).
+2. Add new specialist sub-agents.
+3. Wire Zero into the rest of your toolchain (MCP, skills, hooks, plugins).
 
-- Use the Go version declared in `go.mod`. Do not hardcode a different local
-  toolchain version in scripts or documentation.
-- Use the repository build and release commands (`make` and
-  `go run ./cmd/zero-release ...`) instead of inventing parallel build flows.
-- Keep tests beside their source files (`foo_test.go` next to `foo.go`). Add a
-  regression test for behavior changes and run affected concurrent code under
-  the race detector.
-- Never edit files under `third_party/`; they are vendored.
-- Prefer one cross-platform function with small conditional checks over
-  duplicated platform-specific helpers when the behavior can remain unified.
-- Do not commit generated benchmark reports from
-  `internal/perfbench/reports/*.json`; reports are configuration-specific
-  evidence, not repository state.
-- Preserve the user's working tree. Do not overwrite, delete, stage, or commit
-  unrelated tracked or untracked files.
+## 1. Drop a project `AGENTS.md`
 
-## 3. Required Validation
+When Zero starts in a directory, it looks for project-level instructions and injects them into the system prompt. The lookup walks from your current working directory **up to the nearest git root** and reads the first matching file at each level — general rules at the repo root, more specific rules in sub-trees. Files are labeled with their directory in the prompt (e.g. `## Project guidelines (services/api/AGENTS.md)`).
 
-Run validation from the repository root before committing, opening a pull
-request, or completing an implementation task:
+Accepted file names, in priority order at each level:
 
-1. **Formatting check**: `make fmt-check`. If it fails, format with
-   `go fmt ./...` (or `make fmt`) and run the check again.
-2. **Vet**: `go vet ./...` (or `make vet`).
-3. **Tests**: `go test ./...`. Use `make test` for the full race-enabled suite,
-   or run focused tests with `-race`, when concurrency is affected.
-4. **Build**: `go run ./cmd/zero-release build`.
-5. **Smoke test**: `go run ./cmd/zero-release smoke`.
-6. **Advisory lint**: `make lint-static`.
-7. **Security**: `make vulncheck`.
-8. **Diff hygiene**: `git diff HEAD --check` (covers staged and unstaged
-   tracked changes).
+| Path | Notes |
+| --- | --- |
+| `./AGENTS.md` | The classic spot — committed to your repo, shared with the team. |
+| `./ZERO.md` | Brand-specific alias. Same format, lower priority. |
+| `./.zero/AGENTS.md` | Project-local, hidden, gitignored. Personal notes that stay out of git. |
 
-`make lint` currently runs the formatting check and `go vet`; it does **not**
-run golangci-lint. The pinned golangci-lint job is advisory in CI while the
-existing repository-wide backlog is cleaned up. Fix findings introduced by or
-related to the current change. Unrelated pre-existing advisory findings do not
-justify expanding a focused pull request; report them separately.
+Matching is **case-insensitive** on the basename, so `AGENTS.md`, `Agents.md`, and `agents.md` resolve to the same file on Windows and macOS. The git-tracked filename in this repo is `AGENTS.md` — keep that on case-sensitive filesystems (Linux, the WSL filesystem, or a CI runner) to match what the loader looks for.
 
-Formatting, vet, tests, build, smoke, diff hygiene, and govulncheck are hard
-requirements. If a related check fails, fix it. If a required check cannot run
-because of the environment or fails for an unrelated external reason, report
-the exact failure and obtain maintainer direction rather than silently ignoring
-it.
+Both files use the same format. YAML frontmatter is optional; the markdown body is loaded as instructions for the agent. Zero reads the file once at session start, so changes take effect on the next `zero` launch — not mid-session.
 
-## 4. Common Review Blockers
+```markdown
+# Project conventions for <your project>
 
-These classes drive multi-round reviews. Fix them before requesting review:
+- Build with `make`, not `go build` directly.
+- Tests live next to the source file (`foo_test.go` next to `foo.go`).
+- Run `make lint` before opening a PR.
+- Never edit files under `third_party/` — those are vendored.
+```
 
-- **Fresh base:** Rebase onto the current PR base (`main` or stacked target)
-  before review. A stale head that rolls back mainline commits is a hard
-  blocker, not a merge-time detail. Resolve conflicts by keeping upstream
-  behavior unless the PR intentionally changes it.
-- **Platform truth:** Code and tests must pass on Linux, macOS, and Windows.
-  Compare paths after canonicalization (physical/`EvalSymlinks` form); never
-  assert raw `t.TempDir()` spellings (`/var` vs `/private/var` on macOS, short
-  paths on Windows). Prefer one cross-platform function with small `GOOS`
-  checks over duplicated helpers that drift.
-- **Security edges:** Keep secrets out of argv, env dumps, and logs. Redact
-  success and error paths (including stderr). Fail closed on ownership, lease,
-  and permission checks. Do not rely on pre-open path resolution
-  (`EvalSymlinks` then open) for containment: that is a check-to-use race.
-  Bind containment at open/use time with rooted or handle-relative,
-  traversal-resistant APIs. If a no-follow API is used, apply it to every
-  traversed component and enforce the platform's reparse-point protections;
-  final-component-only no-follow is insufficient. On multi-step setup, roll back
-  only what this run created; never destroy pre-existing resources you did
-  not create; never report success when cleanup or unlock failed.
-- **Atomic shared state:** Write a complete temporary file, then atomically
-  replace the destination so concurrent readers never see a partial write.
-  Exclusive create or a write lock alone is not enough for readers. Serialize
-  the full read-modify-write sequence for lockfiles and shared stores.
-- **Tests match the claim:** Every behavior or security-boundary change needs a
-  regression test, including the failure path. Path-sensitive logic needs at
-  least one non-Linux path case (or a hermetic fake that exercises the same
-  normalization).
-- **Honest scope:** PR description, help text, and comments must match what
-  shipped. Wire advertised entry points or shrink the claim. Do not bundle
-  unrelated fixes. Do not widen security allowlists by name alone; check
-  classification and side effects too.
+Tips:
+
+- Keep each file under ~8 KiB. Zero caps the **total** across all matched files at 32 KiB; everything past the cap is dropped.
+- Re-state rules in the imperative voice: "Run `make lint`", not "you should consider running the linter".
+- Don't put secrets, model IDs, or environment-specific paths in `AGENTS.md`. Use `config.json` for those.
+- In a monorepo, drop a narrower `AGENTS.md` in each sub-tree (e.g. `services/api/AGENTS.md`). Zero picks those up automatically when you launch from inside the sub-tree.
+- A YAML frontmatter block (`---\n...\n---`) at the top is preserved verbatim in the injected prompt but is not parsed for `globs:` or `alwaysApply:` scoping today — keep the body self-contained.
+
+### Personal guidelines, across every project
+
+For preferences that follow *you*, not a specific repo (tone, tooling habits, workflow), drop a `ZERO.md` in your user config directory: `~/.config/zero/ZERO.md` on Linux/macOS, `%AppData%\Roaming\zero\ZERO.md` on Windows — the same directory as `config.json` and your personal specialists. Same format and 8 KiB cap as the project files above, and the same case-insensitive basename match.
+
+This file is injected as its own `## User guidelines` section, before the project's `AGENTS.md`/`ZERO.md`, and is labeled as personal preference in the prompt: project guidelines are the later, more specific instruction and take precedence over it when the two conflict.
+
+## 2. Custom specialists
+
+Specialists are Zero's sub-agents. Three scopes, in priority order:
+
+| Scope | Path | Shared? |
+| --- | --- | --- |
+| Built-in | compiled into Zero | yes — `worker`, `explorer`, `code-review` |
+| User | `~/.config/zero/specialists/*.md` | no — your machine only |
+| Project | `./.zero/specialists/*.md` | yes — the repo team |
+
+Project overrides user overrides built-in when names collide.
+
+A specialist is a markdown manifest with frontmatter and a system prompt:
+
+```markdown
+---
+description: Reviews API changes for breaking-change risk and missing tests.
+tools: read-only,plan
+---
+
+You review API changes. For every changed hunk in `internal/api/` or any file
+that ends in `_api.go`:
+
+1. Confirm the public signature is backward-compatible, or note the breaking
+   change explicitly with the migration path.
+2. Confirm a corresponding test exists in `internal/api/*_test.go` and that
+   the new behaviour is exercised.
+3. Flag any new exported symbol without a doc comment.
+
+Reply with one JSON object per finding: `{"file", "line", "severity", "message", "fix"}`.
+```
+
+CLI management (the prompt is passed inline via `--prompt`):
+
+```bash
+zero specialist list
+zero specialist show api-reviewer
+zero specialist create api-reviewer \
+    --project \
+    --description "Reviews API changes" \
+    --tools read-only,plan \
+    --prompt "$(cat api-reviewer.md)"
+zero specialist edit api-reviewer --project
+zero specialist delete api-reviewer --project
+zero specialist path                       # prints the resolved specialists directory
+```
+
+The full format spec (frontmatter fields, tool scopes, prompt conventions) is in [`docs/SPECIALISTS.md`](docs/SPECIALISTS.md).
+
+> **Roadmap.** An in-UI specialist manager (create / edit / delete / preview) is on the backlog. Today you use the `zero specialist` CLI subcommands above.
+
+## 3. Skills
+
+Skills are markdown instruction packs the agent can pull in on demand. Each skill is a directory containing a `SKILL.md`. Skills are **user-level only** in this version — there's no project-scoped skill directory yet, so anything you want shared with the team goes in `AGENTS.md` (section 1) or as a hook (section 4).
+
+Discovery root: `$ZERO_SKILLS_DIR` → `$XDG_DATA_HOME/zero/skills` → `~/.local/share/zero/skills/`. A missing directory is fine — Zero just reports "no skills".
+
+```
+~/.local/share/zero/skills/
+  run-benchmarks/
+    SKILL.md
+  write-changelog/
+    SKILL.md
+```
+
+`SKILL.md` format:
+
+```markdown
+---
+description: Run the project's benchmark suite and summarize the deltas.
+---
+
+# Run benchmarks
+
+1. `make bench` — captures the wall-clock and RSS before and after.
+2. `benchstat before.txt after.txt` — diffs the two.
+3. Report any regression > 5% with the function name and the previous value.
+```
+
+Only `name` and `description` are recognized in the frontmatter today. The `name` defaults to the directory name. If two skills declare the same name, the one in the lexicographically-first directory wins — duplicates are dropped silently. Plugin-declared skills (section 6) are not yet merged into the loader, so a `skills:` entry inside a plugin's `plugin.json` is not visible to the `skill` tool today.
+
+The `skill` core tool lets the agent load any discovered skill by name.
+
+## 4. Hooks
+
+Hooks fire shell commands on lifecycle events. Configure them in JSON:
+
+- User: `~/.config/zero/hooks.json`
+- Project: `./.zero/hooks.json`
+
+```json
+{
+  "enabled": true,
+  "hooks": [
+    {
+      "id": "block-rm-rf",
+      "event": "beforeTool",
+      "matcher": "bash",
+      "command": "/usr/local/bin/zero-hook-block-rmrf.sh",
+      "enabled": true
+    },
+    {
+      "id": "log-session",
+      "event": "sessionStart",
+      "command": "/usr/local/bin/zero-hook-log.sh",
+      "enabled": true
+    }
+  ]
+}
+```
+
+The `args` array (when present) is passed verbatim to `exec.CommandContext`. The actual hook payload — event name, matcher, tool call id, tool name, tool input, tool output, status — is delivered to the command as **JSON on stdin**, not via `${...}` substitution. A typical handler reads stdin and decides what to do:
+
+```bash
+#!/usr/bin/env bash
+# /usr/local/bin/zero-hook-block-rmrf.sh
+set -euo pipefail
+payload="$(cat)"
+if printf '%s' "$payload" | grep -q '"input":"[^"]*rm[[:space:]]+-rf'; then
+  echo "refusing rm -rf" >&2
+  exit 1
+fi
+```
+
+Events the agent emits (in dispatch order):
+
+| Event | Fires when | Matcher allowed? |
+| --- | --- | --- |
+| `beforeTool` | A tool is about to run | yes (tool name) |
+| `afterTool` | A tool just returned | yes (tool name) |
+| `sessionStart` | A session begins | no |
+| `sessionEnd` | A session ends | no |
+| `specialistStart` | A sub-agent is spawned | yes (specialist name) |
+| `specialistStop` | A sub-agent ends | yes (specialist name) |
+
+A hook's exit code decides what happens next: `0` continues, non-zero blocks the tool call (`beforeTool`) or surfaces an error (`afterTool`). Hook execution is recorded in the audit log; the audit is reachable from the agent's view of past actions, not from a dedicated `zero doctor` check.
+
+> **Roadmap.** An in-UI hooks manager is on the backlog. Today you edit the JSON directly.
+
+## 5. MCP — Model Context Protocol
+
+Zero is both an **MCP client** (it can call external MCP servers) and an **MCP server** (other agents can call its tools).
+
+### As a client — configure MCP servers in `config.json`
+
+```json
+{
+  "mcp": {
+    "servers": {
+      "docs": {
+        "type": "stdio",
+        "command": "docs-mcp",
+        "args": ["--port", "7777"]
+      },
+      "github": {
+        "type": "http",
+        "url": "https://api.example.com/mcp",
+        "headers": { "Authorization": "Bearer YOUR_TOKEN_HERE" }
+      }
+    }
+  }
+}
+```
+
+Manage via CLI:
+
+```bash
+zero mcp add docs --type stdio -- docs-mcp --port 7777
+zero mcp add github --type http --url https://api.example.com/mcp \
+    --header "Authorization=Bearer YOUR_TOKEN_HERE"
+zero mcp list
+zero mcp check docs
+zero mcp remove github
+zero mcp oauth login github
+```
+
+Servers are merged from user and project configs (project wins on conflicts). Token-bearing values in `config.json` are sent verbatim — there is no `${env:...}` expansion — so prefer one of:
+
+- A wrapper script that sources the secret and execs the real command.
+- A `--header` value produced by command substitution (`"Authorization=Bearer $(print-token)"`) in a private shell config that you keep out of git.
+- A secret manager that injects the env var your MCP server reads on its own (the `command` and `args` then run inside that environment).
+
+### As a server — expose Zero's tools to another agent
+
+```bash
+zero serve --mcp
+```
+
+The server speaks MCP over stdio. Configure it from the receiving side as a `stdio` server whose command is `zero serve --mcp`.
+
+## 6. Plugins
+
+A plugin is a self-contained directory that bundles tools, hooks, and skills for one capability. Plugins live at:
+
+- User: `~/.config/zero/plugins/<id>/`
+- Project: `./.zero/plugins/<id>/`
+
+Each plugin has a `plugin.json` manifest:
+
+```json
+{
+  "id": "github-pr-review",
+  "name": "GitHub PR Review",
+  "description": "Adds review tools for GitHub PRs.",
+  "version": "1.0.0",
+  "tools": [
+    { "name": "list_prs", "command": "./tools/list_prs.sh" }
+  ],
+  "hooks": [
+    { "name": "pre-merge-check", "event": "beforeTool", "command": "./hooks/pre-merge.sh" }
+  ],
+  "skills": [
+    { "path": "./skills/review-checklist/SKILL.md" }
+  ]
+}
+```
+
+Install and manage:
+
+```bash
+zero plugins add ./github-pr-review      # copy into ~/.config/zero/plugins/ or ./.zero/plugins/
+zero plugins list
+zero plugins remove github-pr-review    # alias: rm
+```
+
+A plugin is enabled by being present in the plugins directory and disabled by removing it (or by the user setting `"enabled": false` in its `plugin.json`). Plugins are not enabled or disabled by a CLI subcommand today.
+
+Plugin commands run with the plugin directory as their working directory. Use relative paths; the loader resolves them at activation time.
+
+> **Roadmap.** An in-UI plugins manager (browse, install, enable / disable) is on the backlog. Today you use the `zero plugins` CLI subcommands above. Skills declared inside a plugin's `plugin.json` are not yet merged into the `skill` tool's discovery (see section 3).
+
+## 7. Configuration locations
+
+Three layers, applied in order (later layers override earlier ones):
+
+| Layer | Path | Notes |
+| --- | --- | --- |
+| Built-in defaults | compiled in | Lowest priority. |
+| User config | `~/.config/zero/config.json` | Your machine. Never committed. |
+| Project config | `./.zero/config.json` | The repo. Committed (or not, your call). |
+| CLI flags | `--model`, `--mode`, ... | Highest priority, per-invocation. |
+| Environment | `ZERO_*` | Provider commands, secrets, skills dir override. |
+
+The user config holds things that should follow the user across projects (default provider, default model, theme). The project config holds things the team agreed on (provider catalog, sandbox policies, model restrictions).
+
+The sandbox `additionalWriteRoots` key is **ignored in project config** by design — a checked-out repo cannot widen its own sandbox. Set it in the user config or pass `--add-dir` per-invocation.
+
+## 8. End-to-end example
+
+A team that wants every contributor's Zero to behave the same way commits:
+
+- `AGENTS.md` — project conventions, build commands, do-not-edit lists.
+- `.zero/config.json` — provider catalog, default model, allowed tools.
+- `.zero/specialists/api-reviewer.md` — the team's PR-review specialist.
+- `.zero/hooks.json` — block `rm -rf` and `git push --force` on `beforeTool`.
+- `.zero/plugins/internal-tooling/` — a plugin that adds the team's internal CLI tools to the agent's toolset.
+
+Each contributor adds only:
+
+- `~/.config/zero/config.json` — their personal API keys, theme, default mode.
+- `~/.config/zero/ZERO.md` — personal preferences that follow them across every project (see section 1).
+- `~/.local/share/zero/skills/` — personal skills they keep across projects.
+
+That's it. Run `zero` from the repo root and the agent has the team's full instruction set, every contributor's personal setup, and nothing else.
+
+## 9. Reference
+
+- [README](README.md) — install, quickstart, command reference.
+- [docs/SPECIALISTS.md](docs/SPECIALISTS.md) — full specialist manifest spec.
+- [docs/STREAM_JSON_PROTOCOL.md](docs/STREAM_JSON_PROTOCOL.md) — `zero exec` I/O contract.
+- [docs/INSTALL.md](docs/INSTALL.md) — install from source or release.

CHANGELOG.md

@@ -6,279 +6,6 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
 aims to follow [Semantic Versioning](https://semver.org/spec/v2.0.0.html) once the first release is
 tagged. Until then, source builds report the version `dev`.
 
-## [0.7.0](https://github.com/Gitlawb/zero/compare/v0.6.0...v0.7.0) (2026-08-10)
-
-
-### Features
-
-* **agent:** add PermissionModePlan for interactive read-only planning ([#853](https://github.com/Gitlawb/zero/issues/853)) ([0176ee4](https://github.com/Gitlawb/zero/commit/0176ee49ffddcff6f7edf6ae2ade4967e5b22cc3))
-* **providers:** live model lists for OpenRouter and OpenGateway ([#860](https://github.com/Gitlawb/zero/issues/860)) ([fb4f293](https://github.com/Gitlawb/zero/commit/fb4f293c20eb860ca388a10d22fd566ce282ba1f))
-* **tools:** give tool results an image channel and add view_image ([#843](https://github.com/Gitlawb/zero/issues/843)) ([cd0eb19](https://github.com/Gitlawb/zero/commit/cd0eb194e4df4b844785796c0701216740a56344))
-* **tui:** replace /retitle with local /rename ([#826](https://github.com/Gitlawb/zero/issues/826)) ([5886354](https://github.com/Gitlawb/zero/commit/588635487f22ef4d62d350140a219b7b8bb24119))
-
-
-### Bug Fixes
-
-* **agent:** end the additional_permissions retry loop ([#864](https://github.com/Gitlawb/zero/issues/864)) ([4ca4fa7](https://github.com/Gitlawb/zero/commit/4ca4fa75d31b139dbe55a3badf17da86892ab001))
-* **agent:** tell the user a prefix approval leaves the sandbox ([#885](https://github.com/Gitlawb/zero/issues/885)) ([2ec5c3a](https://github.com/Gitlawb/zero/commit/2ec5c3ae6b8fb543887bdeafdb34c4da057b8d36))
-* **cli:** report MCP servers that failed to start ([#822](https://github.com/Gitlawb/zero/issues/822)) ([#827](https://github.com/Gitlawb/zero/issues/827)) ([021281e](https://github.com/Gitlawb/zero/commit/021281ebbab3f2bbe6a77b2dab4bf121bcfe6b52))
-* **lsp:** dispatch notifications off read loop ([#759](https://github.com/Gitlawb/zero/issues/759)) ([936aa4e](https://github.com/Gitlawb/zero/commit/936aa4e77cf49738560a89395b8ba4d22263a396))
-* **mcp:** name the non-text blocks a tool result drops ([#874](https://github.com/Gitlawb/zero/issues/874)) ([027a670](https://github.com/Gitlawb/zero/commit/027a67080bea511b876c284bc911ef011eceb308))
-* **sandbox:** deny reads of Zero credential stores ([#681](https://github.com/Gitlawb/zero/issues/681)) ([55fa05f](https://github.com/Gitlawb/zero/commit/55fa05f0ab5a01b495345d43464606cd19a31b4f))
-* **sandbox:** stop the Windows write jail honouring Everyone-granted paths ([#865](https://github.com/Gitlawb/zero/issues/865)) ([91b413c](https://github.com/Gitlawb/zero/commit/91b413c514cc1607d5b4e7936780d40569239814))
-* **secrets,worktrees:** fix secret redaction leakage and prune stale worktrees ([#855](https://github.com/Gitlawb/zero/issues/855)) ([ff62f73](https://github.com/Gitlawb/zero/commit/ff62f73b7f4dc5e2472db21acba43cf6140c06d1))
-* **specialist:** make overwrites atomic ([#757](https://github.com/Gitlawb/zero/issues/757)) ([30f6d8f](https://github.com/Gitlawb/zero/commit/30f6d8f591a0b81638b4182ff56c86c8a50a5c70))
-* suppress gosec G204 warning for trusted local config commands ([#848](https://github.com/Gitlawb/zero/issues/848)) ([c52db05](https://github.com/Gitlawb/zero/commit/c52db055b0b213758a33c5c5627a411e052f8de4))
-* **tui:** a copied NUL byte no longer panics the TUI on Windows ([#876](https://github.com/Gitlawb/zero/issues/876)) ([cae0269](https://github.com/Gitlawb/zero/commit/cae0269e77e7a0a493cf799b41c653c9ff9a19de)), closes [#875](https://github.com/Gitlawb/zero/issues/875)
-* **tui:** exclude permission waits from turn time ([#820](https://github.com/Gitlawb/zero/issues/820)) ([15eb16f](https://github.com/Gitlawb/zero/commit/15eb16f736f903c94203207032f98aacbcf06318))
-* **tui:** fall back to CellMotion where AllMotion is unreliable ([#872](https://github.com/Gitlawb/zero/issues/872)) ([7fbd864](https://github.com/Gitlawb/zero/commit/7fbd864e91433441489902571ea394ebc5b7bd0d))
-* **tui:** fill alt-screen frame with theme panel background ([#850](https://github.com/Gitlawb/zero/issues/850)) ([020c28e](https://github.com/Gitlawb/zero/commit/020c28eaf95e2f38853f99fc06a13cd1568bf3c9))
-* **tui:** hide unavailable sidebar shortcut ([#821](https://github.com/Gitlawb/zero/issues/821)) ([845df6c](https://github.com/Gitlawb/zero/commit/845df6c7d28e02f475e90acfe838733eb37e6aa2))
-* use OAuth credentials in provider health checks ([#828](https://github.com/Gitlawb/zero/issues/828)) ([d37de92](https://github.com/Gitlawb/zero/commit/d37de9214b1ba86542f4f5ef3b060bb96849763b))
-
-
-### Performance Improvements
-
-* compact common test output ([#846](https://github.com/Gitlawb/zero/issues/846)) ([edf660a](https://github.com/Gitlawb/zero/commit/edf660ab85131f90a452cee51dc2c3a5cf62d632))
-* improve harness context efficiency ([#838](https://github.com/Gitlawb/zero/issues/838)) ([8e26679](https://github.com/Gitlawb/zero/commit/8e266797a8305e689806eadcf9698da2796481ab))
-
-## [0.6.0](https://github.com/Gitlawb/zero/compare/v0.5.0...v0.6.0) (2026-07-29)
-
-
-### Features
-
-* add persistent session goals ([#803](https://github.com/Gitlawb/zero/issues/803)) ([5d1869e](https://github.com/Gitlawb/zero/commit/5d1869e7bc193c97e07418b57b823905c1151d5d))
-* **providers:** add Fireworks AI provider preset ([#772](https://github.com/Gitlawb/zero/issues/772)) ([58daea7](https://github.com/Gitlawb/zero/commit/58daea7a482927e5ab35afaf97761c60bb35e113))
-* **skills:** report hash drift in skill info ([#795](https://github.com/Gitlawb/zero/issues/795)) ([9fbdc34](https://github.com/Gitlawb/zero/commit/9fbdc341551601c13434cf7b9b6f5e4ea4c483b1))
-
-
-### Bug Fixes
-
-* **acp:** interrupt idle reads on cancellation ([#782](https://github.com/Gitlawb/zero/issues/782)) ([15b7a54](https://github.com/Gitlawb/zero/commit/15b7a5475cfeefddbc59634551451a8872b996b7))
-* **cli:** include plugins info in shell completions ([#794](https://github.com/Gitlawb/zero/issues/794)) ([fa8734f](https://github.com/Gitlawb/zero/commit/fa8734f17927de8832dfe041e043cfb86f93d12a))
-* **config:** kill provider-command process tree via job object on Windows ([#690](https://github.com/Gitlawb/zero/issues/690)) ([5c58256](https://github.com/Gitlawb/zero/commit/5c58256abc29c185b449f8ce79647eb3fae61882))
-* **config:** make the provider-command timeout an upper bound ([#811](https://github.com/Gitlawb/zero/issues/811)) ([#813](https://github.com/Gitlawb/zero/issues/813)) ([1c38e8c](https://github.com/Gitlawb/zero/commit/1c38e8cba50db255f3cda8b7cf4298712b2933e1))
-* **config:** stop billing provider-command fixtures to the 5s timeout ([#810](https://github.com/Gitlawb/zero/issues/810)) ([f5fa4ff](https://github.com/Gitlawb/zero/commit/f5fa4ffc64537897aa78a56edca36d8f1dacd021))
-* **cron:** reserve job IDs atomically ([#686](https://github.com/Gitlawb/zero/issues/686)) ([d9b882e](https://github.com/Gitlawb/zero/commit/d9b882eb0a94f38de8447cb54b57860ccd8ed053))
-* **lint:** resolve static analysis baseline ([#769](https://github.com/Gitlawb/zero/issues/769)) ([583e653](https://github.com/Gitlawb/zero/commit/583e6530bd849a8efcb2b88bbe8c297e8592e62c))
-* **providers:** retry provably pre-send transport failures ([#447](https://github.com/Gitlawb/zero/issues/447)) ([#750](https://github.com/Gitlawb/zero/issues/750)) ([8acd42a](https://github.com/Gitlawb/zero/commit/8acd42a3bd743a35807d762f23a33fb61c171839))
-* **sandbox:** add platform-specific write lock around grant state file ([#752](https://github.com/Gitlawb/zero/issues/752)) ([#755](https://github.com/Gitlawb/zero/issues/755)) ([88fb7b6](https://github.com/Gitlawb/zero/commit/88fb7b6192c1ac8173089967d5e9119f92d7a101))
-* **sandbox:** clarify scoped permission prompts ([#798](https://github.com/Gitlawb/zero/issues/798)) ([9fe02ae](https://github.com/Gitlawb/zero/commit/9fe02ae52e3234cb9dae76e0c5594a287ccbb52a))
-* **sandbox:** deny reads of git's credential stores ([#815](https://github.com/Gitlawb/zero/issues/815)) ([#816](https://github.com/Gitlawb/zero/issues/816)) ([33c94ec](https://github.com/Gitlawb/zero/commit/33c94ec7a3e7987e6e36b5fd5e037ff4e6e84c26))
-* **sandbox:** preserve user config and retry denied writes ([#801](https://github.com/Gitlawb/zero/issues/801)) ([81a5e6c](https://github.com/Gitlawb/zero/commit/81a5e6cb716ba4b7ecf7ff3ff261dfa51fd01098))
-* **sandbox:** scrub the daemon token file pointer ([#677](https://github.com/Gitlawb/zero/issues/677)) ([#818](https://github.com/Gitlawb/zero/issues/818)) ([3e7692d](https://github.com/Gitlawb/zero/commit/3e7692d8f0885d346809869935c81a9cf055d89e))
-* **swarm:** close lifecycle admission on shutdown ([#776](https://github.com/Gitlawb/zero/issues/776)) ([bdbda89](https://github.com/Gitlawb/zero/commit/bdbda89f3e42d93e8058ab3e1e9ef402659530cd))
-* **tui:** fix rendering corruption over multipass + Windows Terminal ([#709](https://github.com/Gitlawb/zero/issues/709)) ([5bb1bc1](https://github.com/Gitlawb/zero/commit/5bb1bc11c3188dcea7709ff07c8b25dda3541de8))
-* **tui:** paste clipboard images with Ctrl+V ([#534](https://github.com/Gitlawb/zero/issues/534)) ([#807](https://github.com/Gitlawb/zero/issues/807)) ([ec3b822](https://github.com/Gitlawb/zero/commit/ec3b8225050eb62f0f41a9582ed2881d2f9dcc8e))
-* **tui:** reset quiet clock for new runs ([#799](https://github.com/Gitlawb/zero/issues/799)) ([a4a805a](https://github.com/Gitlawb/zero/commit/a4a805a6c82c51cb255e909bd87ca183c8805cf2))
-* **windows:** use PowerShell for command execution ([#804](https://github.com/Gitlawb/zero/issues/804)) ([518d2e5](https://github.com/Gitlawb/zero/commit/518d2e53ba3fa073090a9b41c278face85b59b64))
-
-## [0.5.0](https://github.com/Gitlawb/zero/compare/v0.4.0...v0.5.0) (2026-07-22)
-
-
-### Features
-
-* **plugins:** add zero plugins info command ([#773](https://github.com/Gitlawb/zero/issues/773)) ([2479884](https://github.com/Gitlawb/zero/commit/2479884dea49580c853a53536c3bbab22ce8a2bf))
-* **sandbox:** disable the sandbox via config ([#687](https://github.com/Gitlawb/zero/issues/687)) ([#746](https://github.com/Gitlawb/zero/issues/746)) ([a21a052](https://github.com/Gitlawb/zero/commit/a21a052a4a32ce9cb3c93c94350cd24e138532cd))
-* **sandbox:** unify command execution and enforcement ([#781](https://github.com/Gitlawb/zero/issues/781)) ([96859c9](https://github.com/Gitlawb/zero/commit/96859c9bd16f6dad4e332efc7e68178b9116118a))
-* **tui:** add /undo as an alias for /rewind ([#698](https://github.com/Gitlawb/zero/issues/698)) ([#747](https://github.com/Gitlawb/zero/issues/747)) ([8c6d302](https://github.com/Gitlawb/zero/commit/8c6d3022bd7801eaad6a0440bb21a27221887297))
-* **tui:** add isolated /btw conversations ([#748](https://github.com/Gitlawb/zero/issues/748)) ([2e267bd](https://github.com/Gitlawb/zero/commit/2e267bdbee4a77813e93a7ab51f0c575b66cee8c))
-* **tui:** permission prompt takes free-text feedback inline ([#780](https://github.com/Gitlawb/zero/issues/780)) ([3967d49](https://github.com/Gitlawb/zero/commit/3967d49d64998341af8ef6d7523fc63aeb2a5a7a))
-
-
-### Bug Fixes
-
-* **agent:** stop write_stdin session_id probing thrash ([#702](https://github.com/Gitlawb/zero/issues/702)) ([#749](https://github.com/Gitlawb/zero/issues/749)) ([fbf8598](https://github.com/Gitlawb/zero/commit/fbf85984f679058125951fe2d5e4f200e09a3e2f))
-* **cli:** warn when ZERO_PROVIDER overrides a providers-use selection ([#767](https://github.com/Gitlawb/zero/issues/767)) ([3524f79](https://github.com/Gitlawb/zero/commit/3524f795e5fdbb827166a84f03351cedfc9eba30))
-* **doctor:** detect missing native binary during runtime checks ([#450](https://github.com/Gitlawb/zero/issues/450)) ([7796022](https://github.com/Gitlawb/zero/commit/77960229b839dca856616f846839aa773f2923f7))
-* **lsp:** make the real-gopls check opt-in so a broken gopls can't fail the suite ([#684](https://github.com/Gitlawb/zero/issues/684)) ([#766](https://github.com/Gitlawb/zero/issues/766)) ([b1f4173](https://github.com/Gitlawb/zero/commit/b1f41735a7f6b7928a1875e986a4dca39d106cb9))
-* make extension installs transactional ([#762](https://github.com/Gitlawb/zero/issues/762)) ([baa4be1](https://github.com/Gitlawb/zero/commit/baa4be13ac5321da4e9f53e864dd1cd395481200))
-* **oauth:** refuse redirects on credential POSTs ([#729](https://github.com/Gitlawb/zero/issues/729)) ([#741](https://github.com/Gitlawb/zero/issues/741)) ([974fc03](https://github.com/Gitlawb/zero/commit/974fc036c2f9a194722f2e8fddbb4fdbf797effe))
-* **oauth:** validate discovered endpoints before merge/use ([#511](https://github.com/Gitlawb/zero/issues/511)) ([#739](https://github.com/Gitlawb/zero/issues/739)) ([ce4a996](https://github.com/Gitlawb/zero/commit/ce4a996ffac4482e704f0fd61b3e442398fb2401))
-* **perfbench:** absolutize the bench binary and make errored tasks first-class ([#730](https://github.com/Gitlawb/zero/issues/730)) ([dbd9443](https://github.com/Gitlawb/zero/commit/dbd94430143df6754d68551d1028ad8f15b82f1b))
-* **perfbench:** grant write tools so mutating tasks measure real edits ([#763](https://github.com/Gitlawb/zero/issues/763)) ([e1975c1](https://github.com/Gitlawb/zero/commit/e1975c1b396b3236ed648870230b4f85863b1d03))
-* **perfbench:** keep the stamped answer file out of negative oracle greps ([#737](https://github.com/Gitlawb/zero/issues/737)) ([015452c](https://github.com/Gitlawb/zero/commit/015452c1c98a39eabb021324182094e188a8bd47))
-* **providers:** stop "provider not found" for env-derived profiles ([#716](https://github.com/Gitlawb/zero/issues/716)) ([4cbd144](https://github.com/Gitlawb/zero/commit/4cbd144d11e5cb67bc5fea46f5b294562dac7a1a))
-* **sandbox:** AST second opinion for interactive-command bypasses ([#473](https://github.com/Gitlawb/zero/issues/473)) ([#745](https://github.com/Gitlawb/zero/issues/745)) ([f079b90](https://github.com/Gitlawb/zero/commit/f079b90f82dac7b7ae0864b279dc9094e31a6627))
-* **sandbox:** bind Windows elevated ACL setup to one no-follow handle ([#765](https://github.com/Gitlawb/zero/issues/765)) ([4945684](https://github.com/Gitlawb/zero/commit/4945684fa26aa5994eda59dcabedc817423c535d))
-* **sandbox:** don't auto-allow shell when re-entrancy skips wrapping ([#727](https://github.com/Gitlawb/zero/issues/727)) ([#744](https://github.com/Gitlawb/zero/issues/744)) ([6849011](https://github.com/Gitlawb/zero/commit/684901165d1b7f50a8bb4af31b1c1951e6926a79))
-* **tools:** give write_stdin's invalid-session errors the same recovery guidance ([#749](https://github.com/Gitlawb/zero/issues/749) follow-up) ([#768](https://github.com/Gitlawb/zero/issues/768)) ([da9fb50](https://github.com/Gitlawb/zero/commit/da9fb50f549f6701eb136fc99f6178d0772d9334))
-* **tools:** read_file recovers a backwards line range instead of erroring ([#779](https://github.com/Gitlawb/zero/issues/779)) ([89bdc67](https://github.com/Gitlawb/zero/commit/89bdc6719a1e1b3a3ef0e36b91a7839b3efdfba9))
-* **tui:** cache settled alt-screen transcript ([#647](https://github.com/Gitlawb/zero/issues/647)) ([d74ceb1](https://github.com/Gitlawb/zero/commit/d74ceb11271ed68a21c19248210e098f411805fb))
-* **tui:** model rows labelled by id when the description is prose; keep the sidebar under the / palette ([#775](https://github.com/Gitlawb/zero/issues/775)) ([b30c397](https://github.com/Gitlawb/zero/commit/b30c3971b40b06ba52ad649ba9dc0ad560d4be4c))
-* **tui:** stop the permission card clashing on cool themes ([#778](https://github.com/Gitlawb/zero/issues/778)) ([722bb31](https://github.com/Gitlawb/zero/commit/722bb3121682d9cd9cd4bc6c127e9014d719262a))
-
-
-### Performance Improvements
-
-* **agent:** concurrent read-only tool batches via capability gate ([#715](https://github.com/Gitlawb/zero/issues/715)) ([31d45d5](https://github.com/Gitlawb/zero/commit/31d45d5f14e915acb9946e8b8eb81632c48f126a))
-* **agent:** execution profiles with one-shot posture escalation (PR10b+PR10c) ([#740](https://github.com/Gitlawb/zero/issues/740)) ([378d538](https://github.com/Gitlawb/zero/commit/378d538e240c289e57821e9f9628f76034419d01))
-* **agent:** posture-escalation signals and controller (PR10a) ([#736](https://github.com/Gitlawb/zero/issues/736)) ([af875df](https://github.com/Gitlawb/zero/commit/af875df58775484304fc65586bd6c74552ad01a2))
-* **agent:** preserve prompt cache prefixes ([#760](https://github.com/Gitlawb/zero/issues/760)) ([739a47e](https://github.com/Gitlawb/zero/commit/739a47e3eac92c3decc8734f52a4d99c7480c3ca))
-* **openai:** optimized turn session — background prewarm and prefix telemetry (PR8) ([#723](https://github.com/Gitlawb/zero/issues/723)) ([60dc84e](https://github.com/Gitlawb/zero/commit/60dc84e7a38c5544ebc047f3cfaf4625dd1e83b5))
-* **output:** add token-aware semantic output budgeting (PR11) ([#717](https://github.com/Gitlawb/zero/issues/717)) ([e5670c4](https://github.com/Gitlawb/zero/commit/e5670c427ff39c628fb0822fd5c9317ee2174583))
-* **providers:** provider capabilities and default turn-session adapter (PR7) ([#720](https://github.com/Gitlawb/zero/issues/720)) ([30e2c3f](https://github.com/Gitlawb/zero/commit/30e2c3f7ffa1d5e487bd10b59d4e823cda191d48))
-* **turn-bench:** Phase 0 — strengthen oracles so pass rate can't be misread as correctness ([#712](https://github.com/Gitlawb/zero/issues/712)) ([727ad4d](https://github.com/Gitlawb/zero/commit/727ad4d321fab45d0cf40f8535522e3d94e55c4a))
-
-## [0.4.0](https://github.com/Gitlawb/zero/compare/v0.3.0...v0.4.0) (2026-07-17)
-
-
-### Features
-
-* **aimlapi:** AI/ML API provider with guided onboarding (top-up + key issuance) ([#655](https://github.com/Gitlawb/zero/issues/655)) ([6b9c2f0](https://github.com/Gitlawb/zero/commit/6b9c2f0c083e02bcd3aad68fb7af0501a9e7cd61))
-* **cli:** show ZERO wordmark on --version ([#673](https://github.com/Gitlawb/zero/issues/673)) ([9acb411](https://github.com/Gitlawb/zero/commit/9acb4113cc3337b3f361a16278e9cf11ca105e34))
-* **cli:** wire MCP serve WorkspaceRoot and --add-dir scope ([#694](https://github.com/Gitlawb/zero/issues/694)) ([75a78e7](https://github.com/Gitlawb/zero/commit/75a78e715bf23154f69c8c79cb58eb4b535b2a2a))
-* **npm:** ship the native binary as platform optionalDependencies ([#626](https://github.com/Gitlawb/zero/issues/626)) ([5e1405d](https://github.com/Gitlawb/zero/commit/5e1405d0b7abff5b3ccb3cfdb66d64d6d3322922))
-* **perf:** emit prompt-prefix hash fingerprint per turn ([#704](https://github.com/Gitlawb/zero/issues/704)) ([1c5c6e7](https://github.com/Gitlawb/zero/commit/1c5c6e78a8a0e228bdf53d7be90934fbce9d98c3))
-* **providers:** add AI/ML API preset (rebased onto main) ([#621](https://github.com/Gitlawb/zero/issues/621)) ([d66a9dd](https://github.com/Gitlawb/zero/commit/d66a9dda69c32aa59f4bea903cfefe00d4b7adef))
-* **providers:** refresh MiniMax model coverage ([#665](https://github.com/Gitlawb/zero/issues/665)) ([fa3052a](https://github.com/Gitlawb/zero/commit/fa3052a1422a4ad30a3a6295829564f42dee31a8))
-* **skills:** discover shared ~/.agents/skills with multi-root skill loading ([#696](https://github.com/Gitlawb/zero/issues/696)) ([7d57999](https://github.com/Gitlawb/zero/commit/7d579996b43741a2e57e3e38fcdd8484fcfc34e9))
-* **tui:** Ctrl+X leader chords and emacs menu navigation ([#699](https://github.com/Gitlawb/zero/issues/699)) ([7f669f4](https://github.com/Gitlawb/zero/commit/7f669f455021be51319ee3b8298cd48a17f745c7))
-* **tui:** press up to edit queued messages ([#656](https://github.com/Gitlawb/zero/issues/656)) ([4c986d3](https://github.com/Gitlawb/zero/commit/4c986d327b5ed26338d295c23bea5839681db8d0))
-
-
-### Bug Fixes
-
-* **acp:** make truncateHint rune-safe ([#614](https://github.com/Gitlawb/zero/issues/614)) ([ddc4927](https://github.com/Gitlawb/zero/commit/ddc4927aac5544bf4dd2c46615aeda5a81c96576))
-* **agent,tui:** resolve git branch detection when starting Zero in subdirectories ([#613](https://github.com/Gitlawb/zero/issues/613)) ([0184581](https://github.com/Gitlawb/zero/commit/0184581ec234ea414e0a47bc33e8b0f4ddfb497b))
-* **agent:** raise default and deep-mode turn budgets ([#650](https://github.com/Gitlawb/zero/issues/650)) ([635c93a](https://github.com/Gitlawb/zero/commit/635c93af51ebc20f3e0917917e55a79edfe27c35))
-* **cli:** prevent consuming positional arguments as flag values ([#619](https://github.com/Gitlawb/zero/issues/619)) ([5b4f48d](https://github.com/Gitlawb/zero/commit/5b4f48d2dcb66402c13bde0c3cfe9c9371da19fb))
-* **config:** enforce MCP trust boundary so project config cannot override user disable ([#609](https://github.com/Gitlawb/zero/issues/609)) ([4d8c31c](https://github.com/Gitlawb/zero/commit/4d8c31cf16a3080344e3be7039408fcae71c075d)), closes [#512](https://github.com/Gitlawb/zero/issues/512)
-* **config:** surface unknown/typo'd config fields instead of silently dropping them ([#645](https://github.com/Gitlawb/zero/issues/645)) ([893b7b4](https://github.com/Gitlawb/zero/commit/893b7b424cc203a2fcf92327a4e25c84286a90e0))
-* **cron:** prevent cron job Mutate from clobbering concurrent updates ([#630](https://github.com/Gitlawb/zero/issues/630)) ([e4bd703](https://github.com/Gitlawb/zero/commit/e4bd703cfb28dab2dfa3c2ddba46237e1bb2e164))
-* **daemon:** handle os.ErrPermission as collision during O_EXCL lock creation ([#616](https://github.com/Gitlawb/zero/issues/616)) ([8ea5384](https://github.com/Gitlawb/zero/commit/8ea53841a5d54c37a153779ae86bea010659433c))
-* **exec:** stop false INCOMPLETE downgrades on conversational final messages ([#608](https://github.com/Gitlawb/zero/issues/608)) ([b6117af](https://github.com/Gitlawb/zero/commit/b6117af86d6bc87a4ee66910e99d76cb16b03fed))
-* harden MCP credential boundaries ([#597](https://github.com/Gitlawb/zero/issues/597)) ([fdddb05](https://github.com/Gitlawb/zero/commit/fdddb05ba84b1600ae6c3a20028bf83afe474c44))
-* **hooks:** fail closed on launch failures for beforeTool hooks ([#629](https://github.com/Gitlawb/zero/issues/629)) ([dc06fe7](https://github.com/Gitlawb/zero/commit/dc06fe72caf45f72d2cba1e8a835c0f5b405c1e8))
-* **hooks:** run sessionEnd hooks after Esc/Ctrl+C interrupts ([#606](https://github.com/Gitlawb/zero/issues/606)) ([824ecdb](https://github.com/Gitlawb/zero/commit/824ecdbcf9c467c35ef4e2666770fdadcb5bf402))
-* **keyring:** pass generic password via stdin on macOS ([#574](https://github.com/Gitlawb/zero/issues/574)) ([91ea6de](https://github.com/Gitlawb/zero/commit/91ea6ded7503538834a84d090f78670a363c62d3))
-* **lock:** prevent POSIX lock file overwrite and leak on Windows/Unix ([#628](https://github.com/Gitlawb/zero/issues/628)) ([da41c3a](https://github.com/Gitlawb/zero/commit/da41c3a75b782d6e0836fe13346321e40a90fbb4))
-* **openai:** omit prompt_cache_key for openai-compatible providers ([#636](https://github.com/Gitlawb/zero/issues/636)) ([1af5882](https://github.com/Gitlawb/zero/commit/1af58828eb3c22567599c000736c913a290959d2)), closes [#624](https://github.com/Gitlawb/zero/issues/624)
-* **plugins:** resolve relative executable paths against plugin root ([#627](https://github.com/Gitlawb/zero/issues/627)) ([2efe6d5](https://github.com/Gitlawb/zero/commit/2efe6d539e29374b3ef39c2290bdea81f33a228b))
-* **sandbox:** remove windowsWriteRestricted flag to fix DenyRead bypass ([#612](https://github.com/Gitlawb/zero/issues/612)) ([3d96ac7](https://github.com/Gitlawb/zero/commit/3d96ac7e55c760a97f28c0e6ceaf1ec3b4ab717a))
-* **sandbox:** scrub dynamic credential env vars ([#682](https://github.com/Gitlawb/zero/issues/682)) ([9043bae](https://github.com/Gitlawb/zero/commit/9043baedcff7776c7373645b57563cac06b31847))
-* **sandbox:** scrub sensitive credentials from sandbox environment ([#660](https://github.com/Gitlawb/zero/issues/660)) ([6fc1220](https://github.com/Gitlawb/zero/commit/6fc1220f6ac66fb3ae67b637cbbed7068d2213c0))
-* **sandbox:** unblock git fetch/commit/add under the write-restricted sandbox ([#654](https://github.com/Gitlawb/zero/issues/654)) ([5c4815a](https://github.com/Gitlawb/zero/commit/5c4815a66ed07d9cf90b825adfd936d3ac07639d))
-* **sandbox:** use WRITE_RESTRICTED token when no DenyRead paths are configured ([#658](https://github.com/Gitlawb/zero/issues/658)) ([a5d2e32](https://github.com/Gitlawb/zero/commit/a5d2e327c8681671aa8a9e5378801215b747edcf))
-* **securefile,credstore:** call Sync on temp file before close and rename ([#631](https://github.com/Gitlawb/zero/issues/631)) ([212734a](https://github.com/Gitlawb/zero/commit/212734adf3b5f982e22385161205aa1afe4634fe))
-* **securefile:** reclaim stale lock files to prevent permanent DOS ([#615](https://github.com/Gitlawb/zero/issues/615)) ([8536cc8](https://github.com/Gitlawb/zero/commit/8536cc87f7a885f9e436d6ef28f5f325201623dc))
-* **swarm:** wait for job.Runs directly in scheduler skip test ([#667](https://github.com/Gitlawb/zero/issues/667)) ([1bb6b57](https://github.com/Gitlawb/zero/commit/1bb6b5745af90321d1e657a12a1976cded5dd1bd))
-* **tools:** classify silent wrapped Windows command failures as sandbox denials ([#659](https://github.com/Gitlawb/zero/issues/659)) ([8bd9742](https://github.com/Gitlawb/zero/commit/8bd9742fa95c41b93ea4e718628aed0ff3ae9dd0))
-* **tools:** preserve SysProcAttr during PTY fallback ([#618](https://github.com/Gitlawb/zero/issues/618)) ([f78b36c](https://github.com/Gitlawb/zero/commit/f78b36c770daa4577a9f99265b18a354454e36eb))
-* **tui:** resolve pending askUser callbacks to prevent runner hangs ([#620](https://github.com/Gitlawb/zero/issues/620)) ([aa73a76](https://github.com/Gitlawb/zero/commit/aa73a76f1bd1b6fe97bac2fbff2d61b7474139f2))
-* **tui:** stop the composer cursor blinking while typing or unfocused ([#672](https://github.com/Gitlawb/zero/issues/672)) ([2b42cd5](https://github.com/Gitlawb/zero/commit/2b42cd567b96d6f7e0818594e53ff31cce1e42e9))
-* **windows:** resolve absolute path for taskkill to prevent hijacking ([#617](https://github.com/Gitlawb/zero/issues/617)) ([2db00ee](https://github.com/Gitlawb/zero/commit/2db00ee3d57db97e0fbe23cb8628e8bcb47f6f09))
-
-
-### Performance Improvements
-
-* **tools:** add explicit effect metadata for safe concurrency ([#705](https://github.com/Gitlawb/zero/issues/705)) ([8ef8576](https://github.com/Gitlawb/zero/commit/8ef8576df7d0775a5b815bc0776a794cadb75c34))
-
-## [0.3.0](https://github.com/Gitlawb/zero/compare/v0.2.0...v0.3.0) (2026-07-09)
-
-
-### Features
-
-* gate project-scoped hooks, plugins, and MCP servers behind workspace trust ([#529](https://github.com/Gitlawb/zero/issues/529)) ([a880ce8](https://github.com/Gitlawb/zero/commit/a880ce80a6ec72da511fb9bdf6dd69291c72a64b))
-* **modelregistry:** infer reasoning efforts for Hunyuan and vendor-prefixed model ids ([#599](https://github.com/Gitlawb/zero/issues/599)) ([92a92ce](https://github.com/Gitlawb/zero/commit/92a92ceb29dc63f5216ab140ef9a6dd9afe17df8))
-* **providers:** OAuth login profiles and list-first /provider manager ([#560](https://github.com/Gitlawb/zero/issues/560)) ([1655056](https://github.com/Gitlawb/zero/commit/16550569c1be615cfaf244dab25909ec37f6dee6))
-* **tui:** remember recent provider+model selections in /model picker ([#568](https://github.com/Gitlawb/zero/issues/568)) ([d0c4e62](https://github.com/Gitlawb/zero/commit/d0c4e62cd429a0614c15296934c740a08bc0e07b))
-* **tui:** show CLI version on the startup home screen ([#538](https://github.com/Gitlawb/zero/issues/538)) ([fd69233](https://github.com/Gitlawb/zero/commit/fd69233e334f1823a06b5794085a9255b3abdfa8))
-* voice dictation (speech-to-text) ([#557](https://github.com/Gitlawb/zero/issues/557)) ([87158a1](https://github.com/Gitlawb/zero/commit/87158a1c90b4f91fc5f2bb8178ebaf46d7654680))
-
-
-### Bug Fixes
-
-* address bugs found in a multi-agent codebase audit ([#481](https://github.com/Gitlawb/zero/issues/481)) ([008bc9b](https://github.com/Gitlawb/zero/commit/008bc9b3f3ba13c7d4822b9559b020f381ff555b))
-* **agent:** keep tools exposed for max-turn finalization ([#533](https://github.com/Gitlawb/zero/issues/533)) ([3f0503b](https://github.com/Gitlawb/zero/commit/3f0503bc2312ae29d5ade784d8824dc9a3524958))
-* **auth:** persist OpenRouter API key after login ([#595](https://github.com/Gitlawb/zero/issues/595)) ([2a062aa](https://github.com/Gitlawb/zero/commit/2a062aa4014c6cd5e20a57dad4a685f88966f109))
-* bump Go to 1.26.5 for crypto/tls fix (GO-2026-5856) ([#607](https://github.com/Gitlawb/zero/issues/607)) ([a7cfb99](https://github.com/Gitlawb/zero/commit/a7cfb99fed7b88ebc09a2f251cb82864d3c2cade))
-* gitignore Windows sandbox helpers and npm version marker ([#578](https://github.com/Gitlawb/zero/issues/578)) ([25653f6](https://github.com/Gitlawb/zero/commit/25653f686c016f95b81ceb7ff5d5452d37c4d4f3))
-* **mcp:** silence startup warning for unconfigured default servers ([#563](https://github.com/Gitlawb/zero/issues/563)) ([302f58b](https://github.com/Gitlawb/zero/commit/302f58bb5f2a03ec7230354ed4747e4e55c16c50))
-* **mcp:** skip RFC 8414 discovery when OAuth endpoints are preconfigured ([#586](https://github.com/Gitlawb/zero/issues/586)) ([8a52d98](https://github.com/Gitlawb/zero/commit/8a52d98cad7cd0086dee9aede4ce477e432bd385))
-* **modelregistry:** reject oversized models.dev cache responses ([#602](https://github.com/Gitlawb/zero/issues/602)) ([66a6396](https://github.com/Gitlawb/zero/commit/66a63964149fb2f07e646e5f1987627c5cd9ac28))
-* **provider:** stop dropping custom no-auth providers on restart ([#558](https://github.com/Gitlawb/zero/issues/558)) ([ba99fa8](https://github.com/Gitlawb/zero/commit/ba99fa8d487fb28f2700e0ff10b2a25c75303cf7))
-* **sandbox:** Windows-appropriate suggestions when blocking interactive commands ([#414](https://github.com/Gitlawb/zero/issues/414)) ([ba4c007](https://github.com/Gitlawb/zero/commit/ba4c00755dc1c31b3dcca18e50b20f62c6bf5d1f))
-* **tools:** block MSYS and WSL shells under the Windows sandbox ([#587](https://github.com/Gitlawb/zero/issues/587)) ([0666818](https://github.com/Gitlawb/zero/commit/066681855b80e3baf5d07d7397610b25f724e353))
-* **tools:** platform-specific pager suggestions, quote/caret-safe cd detection ([#543](https://github.com/Gitlawb/zero/issues/543)) ([8b248f4](https://github.com/Gitlawb/zero/commit/8b248f4e1198dc86ab332a697fba4cf520823cbd))
-* **tools:** Windows cmd.exe quoting guidance and clipboard escaping fix ([#468](https://github.com/Gitlawb/zero/issues/468)) ([f10ed0c](https://github.com/Gitlawb/zero/commit/f10ed0c893ce6de08923f143d681ba96f0fcfe3a))
-* **tui:** bypass toggleSidebar and toggleMouse global shortcuts when composer is non-empty ([#576](https://github.com/Gitlawb/zero/issues/576)) ([c7346fb](https://github.com/Gitlawb/zero/commit/c7346fbcf01fe7e70ed2fcfccbf9965e5985727b))
-* **tui:** paste protection for Termux char-by-char paste ([#573](https://github.com/Gitlawb/zero/issues/573)) ([8e9149f](https://github.com/Gitlawb/zero/commit/8e9149f1a296bebebf95cae9dd2a7c5156c9dbb6))
-* **tui:** update picker_test for switchProviderModel 4-value return ([#589](https://github.com/Gitlawb/zero/issues/589)) ([8f15650](https://github.com/Gitlawb/zero/commit/8f156506dc92449bd24caa14e36f28477ba00fff))
-* **update:** clearer error on unsupported release platform (android/termux) ([#603](https://github.com/Gitlawb/zero/issues/603)) ([1fc9b2d](https://github.com/Gitlawb/zero/commit/1fc9b2d25c79e089f34cb7b5b6a7f7c7b8233123))
-* **update:** support safe symlink extraction during updates ([#575](https://github.com/Gitlawb/zero/issues/575)) ([ce9cb91](https://github.com/Gitlawb/zero/commit/ce9cb912958a3d9ae0b052bfebb849c28e0e719b))
-* warn about untracked scratch files left behind after a run ([#571](https://github.com/Gitlawb/zero/issues/571)) ([062328b](https://github.com/Gitlawb/zero/commit/062328b632a2d27353a6d47c522bbd22d7282539))
-
-
-### Performance Improvements
-
-* **grep:** stop content scan after head limit ([#601](https://github.com/Gitlawb/zero/issues/601)) ([8a05e64](https://github.com/Gitlawb/zero/commit/8a05e6486f7a63281b869064db03dfc5531e6a04))
-
-## [0.2.0](https://github.com/Gitlawb/zero/compare/v0.1.0...v0.2.0) (2026-07-06)
-
-
-### Features
-
-* add --auto flag for LLM-generated commit messages ([#423](https://github.com/Gitlawb/zero/issues/423)) ([b0abde7](https://github.com/Gitlawb/zero/commit/b0abde7d0697e808480cd59d69a6f4d0c6320475))
-* add zero changes push and pr subcommands, and extra repo-info metrics ([#391](https://github.com/Gitlawb/zero/issues/391)) ([2312abe](https://github.com/Gitlawb/zero/commit/2312abe5ddd95f4c6ef373cfb61cc03092f48cdd))
-* agent quality, caching, retry, and tooling upgrades ([#506](https://github.com/Gitlawb/zero/issues/506)) ([3c81fea](https://github.com/Gitlawb/zero/commit/3c81fea22873ee3df7fc97b10cb4f77792706c4b))
-* **agent:** curb over-engineering the solution in the editing discipline ([#517](https://github.com/Gitlawb/zero/issues/517)) ([f4c998a](https://github.com/Gitlawb/zero/commit/f4c998ac30c4f07ff313a2d706791e857293be49)), closes [#516](https://github.com/Gitlawb/zero/issues/516)
-* **agent:** inject per-user config.UserConfigDir()/zero/ZERO.md guidelines into system prompt ([#475](https://github.com/Gitlawb/zero/issues/475)) ([7b10aab](https://github.com/Gitlawb/zero/commit/7b10aab74bf14a01166b2cea22deab79bba9850b))
-* **openai:** forward prompt_cache_key for server-side prefix cache routing ([#515](https://github.com/Gitlawb/zero/issues/515)) ([87e7e69](https://github.com/Gitlawb/zero/commit/87e7e69afd18b5539579856f3a61c6a95bc445ae))
-* **providers:** add `zero providers models` to discover a provider's models ([#386](https://github.com/Gitlawb/zero/issues/386)) ([0bc8074](https://github.com/Gitlawb/zero/commit/0bc8074c97b0310e4a9d70c3f967003ee5e8a59f))
-* **providers:** add KiloCode and OpenCode provider support ([#388](https://github.com/Gitlawb/zero/issues/388)) ([b1ccb6d](https://github.com/Gitlawb/zero/commit/b1ccb6d9c1875377f5e5ea81a1304edd1e41ab4f))
-* **providers:** add Meituan LongCat catalog preset ([#424](https://github.com/Gitlawb/zero/issues/424)) ([b4275e3](https://github.com/Gitlawb/zero/commit/b4275e350472b2490212bf814709819d354c1216))
-* **providers:** split minimax zai into intl cn ([#398](https://github.com/Gitlawb/zero/issues/398)) ([aaad4d2](https://github.com/Gitlawb/zero/commit/aaad4d271270f41af837b6f3b60ae80beba0c645))
-* require manual approval before npm publish + drop release-as pin ([#369](https://github.com/Gitlawb/zero/issues/369)) ([bd89a1f](https://github.com/Gitlawb/zero/commit/bd89a1f451643c1b65ec803070abc7b116631ebe))
-* **sandbox:** unelevated Windows fallback tier instead of prompts-only degrade ([#427](https://github.com/Gitlawb/zero/issues/427)) ([b9ddd6f](https://github.com/Gitlawb/zero/commit/b9ddd6f42138312a1fee8d8bb67c46c8eb1dea2f))
-* support shift enter for composer newlines ([#462](https://github.com/Gitlawb/zero/issues/462)) ([daf65e0](https://github.com/Gitlawb/zero/commit/daf65e0af9a040314d4ab337b0ad59c55416b7bc))
-* **tui:** /loop — repeat a prompt or command on an interval or self-paced ([#502](https://github.com/Gitlawb/zero/issues/502)) ([387fe67](https://github.com/Gitlawb/zero/commit/387fe67ee7cd81317c9c969f5906a4437080fea3))
-* **tui:** add search/filter to provider picker in setup wizard ([#400](https://github.com/Gitlawb/zero/issues/400)) ([2fcea71](https://github.com/Gitlawb/zero/commit/2fcea71778d23e050c93409c471aef45b68c1621))
-* **update:** add zero upgrade command to apply self-updates ([#461](https://github.com/Gitlawb/zero/issues/461)) ([5f36349](https://github.com/Gitlawb/zero/commit/5f36349c1884e81fa9bc66bb5fe813b627e897b7))
-
-
-### Bug Fixes
-
-* **action:** keep provider key scoped to zero step ([#448](https://github.com/Gitlawb/zero/issues/448)) ([407a927](https://github.com/Gitlawb/zero/commit/407a92739ff508cba32d2c12b3f36f0efcdd54c3))
-* add android platform support for Termux npm install ([#455](https://github.com/Gitlawb/zero/issues/455)) ([9bd93c6](https://github.com/Gitlawb/zero/commit/9bd93c62f8d57fb74057284aa66a1b6e1429dcdd)), closes [#449](https://github.com/Gitlawb/zero/issues/449)
-* **agent:** reject a malformed additional_permissions payload before prompting ([#453](https://github.com/Gitlawb/zero/issues/453)) ([e4f760e](https://github.com/Gitlawb/zero/commit/e4f760ee8bd57299cd2fcb37e8e23130037c2607))
-* allow non-TLS connections to private-network provider endpoints ([#444](https://github.com/Gitlawb/zero/issues/444)) ([1d86384](https://github.com/Gitlawb/zero/commit/1d8638466ca31517eb9db2b9353d3dce1cbeeabc))
-* **auth:** route zero auth login chatgpt to the dedicated ChatGPT flow ([#443](https://github.com/Gitlawb/zero/issues/443)) ([305a62c](https://github.com/Gitlawb/zero/commit/305a62c954ca6cec00bc58d5398f933415156aff))
-* **config:** fall back to a usable saved provider instead of forcing full re-onboarding ([#410](https://github.com/Gitlawb/zero/issues/410)) ([c60ad87](https://github.com/Gitlawb/zero/commit/c60ad8729f79bb841114d352ee2d2fe29d5d0e41))
-* **config:** let a gateway ANTHROPIC_BASE_URL resolve as anthropic-compatible ([#497](https://github.com/Gitlawb/zero/issues/497)) ([30dd7c3](https://github.com/Gitlawb/zero/commit/30dd7c3112ad22d42fa12b5addd4e38f4beda42a)), closes [#479](https://github.com/Gitlawb/zero/issues/479)
-* **config:** unbrick first-run setup — default google/anthropic models, enter setup on fixable config errors ([#385](https://github.com/Gitlawb/zero/issues/385)) ([72eed06](https://github.com/Gitlawb/zero/commit/72eed06b4f94c43d75d31fe54a58d2f566de059e))
-* **config:** use ~/.config on macOS and enter setup when no provider ([#371](https://github.com/Gitlawb/zero/issues/371)) ([#372](https://github.com/Gitlawb/zero/issues/372)) ([027a8f2](https://github.com/Gitlawb/zero/commit/027a8f2768b17b89f5c8270887f156e2ccda69ea))
-* **docs:** rename AGENTS.MD > AGENTS.md ([#438](https://github.com/Gitlawb/zero/issues/438)) ([4266baf](https://github.com/Gitlawb/zero/commit/4266baf222df583ed2078b776687f12d496475b5))
-* **gemini:** strip unsupported JSON Schema fields from tool declarations ([#374](https://github.com/Gitlawb/zero/issues/374)) ([39e7100](https://github.com/Gitlawb/zero/commit/39e7100674150144a1152e3110c64c7cf0321d64)), closes [#373](https://github.com/Gitlawb/zero/issues/373)
-* **install:** persist install dir to user PATH on Windows ([#407](https://github.com/Gitlawb/zero/issues/407)) ([bdb1b0e](https://github.com/Gitlawb/zero/commit/bdb1b0ecd15859b1712a6037d296dace7f9c3c3f))
-* **mcp:** block cross-origin credential redirects ([#396](https://github.com/Gitlawb/zero/issues/396)) ([f915f70](https://github.com/Gitlawb/zero/commit/f915f70e5a3096e2419fa8d961a0f84a626fa4a9))
-* **oauth:** treat Windows ERROR_ACCESS_DENIED as lock contention in createSecretFile ([#445](https://github.com/Gitlawb/zero/issues/445)) ([d05e914](https://github.com/Gitlawb/zero/commit/d05e9148a7f79f67d1d3c31fca2775f21fbd331e))
-* **openai:** handle Ollama reasoning stream deltas ([#486](https://github.com/Gitlawb/zero/issues/486)) ([f6c0606](https://github.com/Gitlawb/zero/commit/f6c060631e18e082dda24cc4dc0903c31c2120d6))
-* preserve conversation context in exec prompts ([#460](https://github.com/Gitlawb/zero/issues/460)) ([949ee43](https://github.com/Gitlawb/zero/commit/949ee43f71e5cb7fab4695c5cb7b442fe4ecfbf7))
-* **provider-wizard:** allow multiple custom OpenAI-compatible providers ([#403](https://github.com/Gitlawb/zero/issues/403)) ([3fbbd28](https://github.com/Gitlawb/zero/commit/3fbbd28e4c586822cc4312c86232d94befe56e87))
-* **sandbox:** fix nested pipe creation under the Windows restricted token ([#456](https://github.com/Gitlawb/zero/issues/456)) ([563a6db](https://github.com/Gitlawb/zero/commit/563a6dbe91e65d5daeefd7626e8a77e30a6d8fb2))
-* **sandbox:** gate /tmp test assertions on GOOS, not path existence ([#426](https://github.com/Gitlawb/zero/issues/426)) ([f653dca](https://github.com/Gitlawb/zero/commit/f653dcac363fb69ad7be5b35e6e0fa6d2bce476d))
-* **sandbox:** self-heal a corrupt unelevated setup marker ([#437](https://github.com/Gitlawb/zero/issues/437)) ([8d0c5fe](https://github.com/Gitlawb/zero/commit/8d0c5feccb8bdbfb015df0508aa6e3bcbd1fd0e8))
-* **specialist:** cap max specialist nesting depth ([#491](https://github.com/Gitlawb/zero/issues/491)) ([177442c](https://github.com/Gitlawb/zero/commit/177442cfe4015bd8df04cc9894f98b468ee796d4))
-* Termux/Android support — PRoot scroll, SIGSYS sandbox, build docs ([#509](https://github.com/Gitlawb/zero/issues/509)) ([0f69d99](https://github.com/Gitlawb/zero/commit/0f69d995e9b586b774f66c066b21abab5e03024a))
-* **tools:** Block MSYS coreutils under Windows sandbox ([#476](https://github.com/Gitlawb/zero/issues/476)) ([81aad58](https://github.com/Gitlawb/zero/commit/81aad58d97839e51c068e2f08907618991fdc3fb))
-* **tools:** CRLF line ending mismatch in edit_file tool on Windows ([#378](https://github.com/Gitlawb/zero/issues/378)) ([33dc7ae](https://github.com/Gitlawb/zero/commit/33dc7ae2cc82c5389675531e1416856dae7151ce))
-* **tools:** fix cmd.exe /S/C corrupting commands with embedded quotes ([#465](https://github.com/Gitlawb/zero/issues/465)) ([190241b](https://github.com/Gitlawb/zero/commit/190241bd593f43211b766e0b13c8e89802d4bb37))
-* **tools:** flag piped POSIX utilities before running on Windows ([#412](https://github.com/Gitlawb/zero/issues/412)) ([5658a36](https://github.com/Gitlawb/zero/commit/5658a366274fc59a9d5336b06a21019c9c25cbf1))
-* **tools:** make grep and glob respect run cancellation ([#464](https://github.com/Gitlawb/zero/issues/464)) ([ba6c026](https://github.com/Gitlawb/zero/commit/ba6c0264697b7d7ed479f6e782fba9700a481e3d))
-* **tools:** require permission before web_search requests ([#382](https://github.com/Gitlawb/zero/issues/382)) ([960db96](https://github.com/Gitlawb/zero/commit/960db9660e4e31dc588fe8f7d6f116ff5e225566))
-* **tui:** compose help overlay through the viewport overlay pipeline ([#421](https://github.com/Gitlawb/zero/issues/421)) ([5b2b4de](https://github.com/Gitlawb/zero/commit/5b2b4dea1aaf9e0f68baa25e97e83296fb17b1a2))
-* **tui:** keep the profile name on /model switch so the stored key resolves ([#441](https://github.com/Gitlawb/zero/issues/441)) ([9134148](https://github.com/Gitlawb/zero/commit/9134148f4df3e4e556fba6c2f8babfdf6fcfeee1)), closes [#440](https://github.com/Gitlawb/zero/issues/440)
-* **tui:** resolve every permission request so the agent can't deadlock ([#397](https://github.com/Gitlawb/zero/issues/397)) ([952788f](https://github.com/Gitlawb/zero/commit/952788f72d32957659fe004521fcc8372b9ba9b4))
-* **tui:** show an M suffix for million-scale token counts ([#457](https://github.com/Gitlawb/zero/issues/457)) ([0562e3b](https://github.com/Gitlawb/zero/commit/0562e3bef7df2328610a48a1e81632a8da4aec64))
-* **tui:** title /model rows by model name, not the catalog description ([#395](https://github.com/Gitlawb/zero/issues/395)) ([cdf9d83](https://github.com/Gitlawb/zero/commit/cdf9d839ae57a729f292f36f7c5b0c67b41b288d))
-
-
-### Performance Improvements
-
-* cache TUI model registry ([#496](https://github.com/Gitlawb/zero/issues/496)) ([e7d88b4](https://github.com/Gitlawb/zero/commit/e7d88b4b518049733da25a8447c00144bd1da518))
-* universal tool-output ceiling with spill + async post-edit diagnostics ([#518](https://github.com/Gitlawb/zero/issues/518)) ([95ccd5b](https://github.com/Gitlawb/zero/commit/95ccd5bc327f6fb464ff0239f7229de789f578dc))
-
 ## 0.1.0 (2026-07-02)
 
 
@@ -295,19 +22,14 @@ tagged. Until then, source builds report the version `dev`.
 ## [Unreleased]
 
 ### Added
-- Shared multi-agent skills discovery: when present, `~/.agents/skills` is searched after the primary
-  Zero skills dir (and before plugin skill roots). `zero skills list` / `info` and the runtime `skill`
-  tool share one multi-root discovery path; install/remove/lock still target only the Zero skills directory.
 - `SECURITY.md` with a private vulnerability-reporting path, `CODE_OF_CONDUCT.md`, this changelog, and
   GitHub issue/PR templates.
 - Interactive `/theme` picker: bare `/theme` opens a popup that live-previews each palette as you move
   and applies on select (Esc reverts).
-- Twelve built-in color themes alongside the `dark`/`light` built-ins — `dracula`, `nord`, `gruvbox`,
-  `tokyo-night`, `catppuccin`, `one-dark`, `solarized-dark`, `rose-pine`, `everforest`,
-  `solarized-light`, `dune`, and `neon` — selectable via `/theme <name>`, `--theme <name>`, or
-  `ZERO_THEME`. Every palette is contrast-audited to WCAG AA, and the new presets are additionally
-  audited after xterm-256 downsampling; see [docs/THEMES.md](docs/THEMES.md). The built-in light
-  theme was reworked for legibility.
+- Ten built-in color themes alongside the `dark`/`light` built-ins — `dracula`, `nord`, `gruvbox`,
+  `tokyo-night`, `catppuccin`, `one-dark`, `solarized-dark`, `rose-pine`, `everforest`, and
+  `solarized-light` — selectable via `/theme <name>`, `--theme <name>`, or `ZERO_THEME`. Every palette
+  is contrast-audited to WCAG AA. The built-in light theme was reworked for legibility.
 - `--theme <name>` flag for the TUI, accepting `auto` or any registered theme (previously only the
   `ZERO_THEME` env var existed).
 - "Accessibility / Appearance" section in the README documenting `NO_COLOR`, `ZERO_THEME`, `/theme`,

CONTRIBUTING.md

@@ -8,13 +8,6 @@ stabilizes. Community participation is still welcome, especially through bug
 reports, feature requests, reproduction details, logs, and discussion on existing
 issues.
 
-Questions and setup help belong in
-[GitHub Discussions](https://github.com/Gitlawb/zero/discussions) (see
-[SUPPORT.md](SUPPORT.md)), not in the issue tracker. Feature ideas start in the
-[Ideas](https://github.com/Gitlawb/zero/discussions/categories/ideas) category;
-an idea that gains traction there is the path to an approved issue and,
-eventually, an accepted PR.
-
 Please read this policy before opening a pull request. Pull requests that do not
 follow it may be closed without review.
 
@@ -40,11 +33,7 @@ unless the issue has been explicitly approved for outside contribution.
 There are three primary ways to contribute:
 
 1. Open a bug report with clear reproduction steps.
-2. Propose a feature in
-   [Discussions Ideas](https://github.com/Gitlawb/zero/discussions/categories/ideas),
-   describing the problem, use case, and expected behavior. A feature-request
-   issue is filed only after the idea gains traction there or a maintainer asks
-   for one.
+2. Open a feature request describing the problem, use case, and expected behavior.
 3. Add useful context to an existing issue, such as logs, screenshots, examples,
    affected versions, or reproduction cases.
 
@@ -57,8 +46,7 @@ All community pull requests require an approved parent issue.
 
 Before starting implementation work:
 
-1. For a bug, open an issue using the bug-report template. For a feature, start
-   in Discussions Ideas; the feature-request issue comes after the discussion.
+1. Open an issue using the appropriate issue template.
 2. Describe the bug, feature request, or functional change clearly.
 3. Wait for the core team to review the issue.
 4. Do not open a pull request unless the issue has the `issue-approved` label.
@@ -67,28 +55,6 @@ An approved issue means only that the specific issue has been accepted for work.
 It does not approve unrelated future work, future pull requests, or future issues
 from the same author. Each issue is reviewed case by case.
 
-## Contribution Volume and Batching
-
-To keep triage and review sustainable while the project stabilizes, please keep
-your contribution flow to a size the core team can realistically work through:
-
-- One issue per bug. File each defect as its own issue with its own
-  reproduction. Please do not bundle several unrelated findings into a single
-  "audit" issue; separate issues are far easier to triage, approve, and close
-  individually.
-- Keep only a few items open at a time. If you already have several issues or
-  pull requests awaiting review, let those move through the queue before opening
-  more. A large simultaneous burst from one contributor slows review for
-  everyone, even when each item is valid.
-- AI-assisted contributions are welcome, held to the same bar. Automated audits
-  and generated patches are fine, but each issue and pull request must be
-  individually verified, scoped, and reproducible before you open it. Please do
-  not post bulk machine-generated findings unfiltered.
-
-These apply to all contributors and are about review bandwidth, not the value of
-any single report. Maintainers may ask you to consolidate, split, or hold
-submissions to keep the queue manageable.
-
 ## Pull Request Requirements
 
 Before opening a community pull request, make sure all of the following are true:
@@ -167,12 +133,7 @@ Clear bug reports are the most useful way to help the project during this phase.
 
 ## Submitting a Feature Request
 
-Feature requests start as a thread in
-[Discussions Ideas](https://github.com/Gitlawb/zero/discussions/categories/ideas),
-not as an issue. The feature-request issue form exists for the step AFTER that:
-when an idea has gained traction in Discussions or a maintainer has asked for an
-issue to track it. Whether in the discussion or the eventual issue, please
-include:
+When filing a feature request, please include:
 
 - The problem you are trying to solve.
 - The behavior or functionality you are requesting.
@@ -195,9 +156,5 @@ contribution phase.
 
 ## Getting Help
 
-General questions, setup help, and "is this a bug or my config" uncertainty all
-go to
-[Discussions Q&A](https://github.com/Gitlawb/zero/discussions/categories/q-a).
-If you are unsure whether a proposed code change is appropriate, raise it in
-Discussions Ideas or on the relevant issue first and ask. Please keep discussion
-focused, specific, and respectful of maintainer time.
+If you are unsure whether a change is appropriate, open an issue first and ask.
+Please keep discussion focused, specific, and respectful of maintainer time.

Makefile

@@ -1,13 +1,7 @@
-# Zero build/test/lint targets. AGENTS.md says to build and run quality checks
-# with `make` — these targets back those instructions.
+# Zero build/test/lint targets. AGENTS.md says "Build with `make`" and "Run `make
+# lint` before opening a PR" — these targets back those instructions.
 .DEFAULT_GOAL := build
-GO_VERSION = $(word 2,$(shell git grep -G -h "^go[[:space:]]" -- go.mod))
-GO_TOOLCHAIN = go$(GO_VERSION)
-DEADCODE_VERSION := v0.46.0
-GOLANGCI_LINT_VERSION := v2.12.2
-GOVULNCHECK_VERSION := v1.3.0
-
-.PHONY: build build-all test test-race vet fmt fmt-check lint lint-static deadcode vulncheck tidy clean baseline help
+.PHONY: build build-all test test-race vet fmt fmt-check lint tidy clean help
 
 # Build the main CLI binary into ./zero.
 build:
@@ -39,22 +33,6 @@ fmt-check:
 # Lint = formatting check + vet (no extra tooling required).
 lint: fmt-check vet
 
-# Versioned tools select the toolchain from their own modules when invoked with
-# package@version. Read this module's go directive directly, without invoking
-# the possibly stale Go toolchain or consulting a multi-module GOWORK. git grep
-# works with both POSIX shells and cmd.exe, including GNU Make 3.81 on macOS.
-# The target-specific export is shell-independent.
-lint-static deadcode vulncheck: export GOTOOLCHAIN = $(GO_TOOLCHAIN)
-
-lint-static:
-	go run github.com/golangci/golangci-lint/v2/cmd/golangci-lint@$(GOLANGCI_LINT_VERSION) run --enable-only unused,ineffassign,staticcheck ./...
-
-deadcode:
-	go run golang.org/x/tools/cmd/deadcode@$(DEADCODE_VERSION) -test=false ./...
-
-vulncheck:
-	go run golang.org/x/vuln/cmd/govulncheck@$(GOVULNCHECK_VERSION) ./...
-
 tidy:
 	go mod tidy
 
@@ -62,19 +40,5 @@ clean:
 	rm -f zero
 	go clean ./...
 
-# Run the per-turn benchmark harness over the checked-in baseline manifest and
-# write the JSON result to internal/perfbench/reports/baseline.json. Requires a
-# built `zero` binary and a model; set ZERO_BENCH_MODEL (required) and
-# ZERO_BENCH_BINARY (defaults to ./zero) to configure the run. The report is
-# machine-specific and regenerated, not hand-edited.
-baseline: build
-	@if [ -z "$(ZERO_BENCH_MODEL)" ]; then echo "Set ZERO_BENCH_MODEL (and optionally ZERO_BENCH_BINARY) before running 'make baseline'"; exit 2; fi
-	@ZERO_BIN="$${ZERO_BENCH_BINARY:-./zero}"; \
-	go run ./cmd/zero-perf-bench turn \
-		--suite internal/perfbench/manifests/baseline.json \
-		--model $(ZERO_BENCH_MODEL) \
-		--binary "$$ZERO_BIN" \
-		--output internal/perfbench/reports/baseline.json
-
 help:
-	@echo "Targets: build (default), build-all, test, test-quick, vet, fmt, fmt-check, lint, lint-static, deadcode, vulncheck, tidy, clean, baseline"
+	@echo "Targets: build (default), build-all, test, test-quick, vet, fmt, fmt-check, lint, tidy, clean"

README.md

@@ -6,9 +6,8 @@
 
 <p align="center">
   <a href="LICENSE"><img alt="license" src="https://img.shields.io/badge/license-MIT-blue"></a>
-  <img alt="Go 1.26.5+" src="https://img.shields.io/badge/Go-1.26.5+-00ADD8?logo=go&logoColor=white">
+  <img alt="Go 1.25+" src="https://img.shields.io/badge/Go-1.25+-00ADD8?logo=go&logoColor=white">
   <img alt="25+ providers" src="https://img.shields.io/badge/providers-25+-34E2EA">
-  <a href="https://discord.gg/CaQDS6wdFn"><img alt="Discord" src="https://img.shields.io/badge/Discord-join-5865F2?logo=discord&logoColor=white"></a>
   <br>
   <strong>English</strong> | <a href="README_ZH.md">中文</a>
 </p>
@@ -50,15 +49,34 @@ npm install -g @gitlawb/zero
 zero
 ```
 
-The npm package is a small wrapper whose platform build (Linux and macOS on
-x64/arm64, Windows on x64 — including the browser/terminal control helpers)
-installs as an optional dependency straight from the npm registry — no install
-scripts, no downloads outside npm. Bun, pnpm, and yarn work the same way with
-no trust or approval steps. Installs that skip optional dependencies
-(`--omit=optional`) still work: the wrapper fetches the binary from the
-matching GitHub Release whenever it is missing. Windows on ARM runs the x64
-build under emulation. See [docs/NPM_PACKAGING.md](docs/NPM_PACKAGING.md) for
-how the package is put together.
+The npm package installs a small wrapper plus the matching Zero binary for your
+platform from GitHub Releases. It supports Linux, macOS, and Windows on x64 and
+arm64.
+
+### Bun
+
+Bun does not run dependency lifecycle scripts by default, so the `postinstall`
+that fetches the Zero binary is skipped and the first run fails with
+`No native binary found next to the npm wrapper`.
+
+The simplest fix is to trust the package after installing, which runs the
+blocked postinstall. This works for project and global installs:
+
+```bash
+# project install
+bun add @gitlawb/zero
+bun pm trust @gitlawb/zero
+
+# global install
+bun add -g @gitlawb/zero
+bun pm -g trust @gitlawb/zero
+```
+
+Alternatives: allow the postinstall up front by adding
+`"trustedDependencies": ["@gitlawb/zero"]` to your project's package.json
+before `bun add`, or run the installer manually
+(`node node_modules/@gitlawb/zero/scripts/postinstall.mjs`) on Bun versions
+that do not have `bun pm trust`.
 
 ### Install scripts
 
@@ -76,7 +94,7 @@ irm https://raw.githubusercontent.com/Gitlawb/zero/main/scripts/install.ps1 | ie
 
 ### From source
 
-Source builds require Go 1.26.5+.
+Source builds require Go 1.25+.
 
 ```bash
 git clone https://github.com/Gitlawb/zero.git
@@ -130,17 +148,7 @@ the key in the wizard:
 export OPENAI_API_KEY=sk-...
 export ANTHROPIC_API_KEY=...
 export GEMINI_API_KEY=...
-export AIMLAPI_API_KEY=...
 export LONGCAT_API_KEY=...
-export FIREWORKS_API_KEY=...
-export MINIMAX_API_KEY=...
-export MINIMAXI_API_KEY=...
-```
-
-To configure AI/ML API directly, run:
-
-```bash
-zero providers setup aimlapi --set-active
 ```
 
 To configure Meituan LongCat (LongCat-2.0) directly, run:
@@ -149,39 +157,6 @@ To configure Meituan LongCat (LongCat-2.0) directly, run:
 zero providers setup longcat --set-active
 ```
 
-To configure Fireworks AI directly, run:
-
-```bash
-zero providers setup fireworks --set-active
-```
-
-MiniMax presets use the Anthropic-compatible endpoints for the global and China
-regions:
-
-```bash
-zero providers add minimax --set-active
-zero providers add minimaxi-cn --set-active
-```
-
-To use the OpenAI-compatible endpoints instead, add a custom compatible profile
-for the required region:
-
-```bash
-zero providers add custom-openai-compatible \
-  --name minimax-openai \
-  --model MiniMax-M3 \
-  --base-url https://api.minimax.io/v1 \
-  --api-key-env MINIMAX_API_KEY \
-  --set-active
-
-zero providers add custom-openai-compatible \
-  --name minimax-cn-openai \
-  --model MiniMax-M3 \
-  --base-url https://api.minimaxi.com/v1 \
-  --api-key-env MINIMAXI_API_KEY \
-  --set-active
-```
-
 For local models, run Ollama or LM Studio and then use `zero setup` or
 `zero providers detect`.
 
@@ -199,11 +174,9 @@ Useful controls:
 |---|---|
 | `Enter` | send the prompt |
 | `/` | open slash-command suggestions |
-| `Ctrl+X` then letter | common slash commands (e.g. `m` → `/model`; `Ctrl+X` `?` for full list) |
-| `Ctrl+P` / `Ctrl+N` | previous / next item in menus (arrows still work) |
 | `Shift+Tab` | cycle permission mode |
 | `Ctrl+B` | show/hide the sidebar |
-| `Ctrl+C` | cancel, exit, or return from a `/btw` conversation |
+| `Ctrl+C` | cancel or exit |
 
 Common slash commands:
 
@@ -213,8 +186,6 @@ Common slash commands:
 | `/spec`, `/plan` | draft and review a plan before building |
 | `/image` | attach an image for vision-capable models |
 | `/resume`, `/rewind` | continue or roll back local sessions |
-| `/new` | start a fresh session in place (previous session stays on disk) |
-| `/btw [question]` | ask in an isolated fork without adding the side conversation to the main session |
 | `/loop` | repeat a prompt or custom `/command` on an interval (`/loop 5m /babysit-prs`) or self-paced |
 | `/compact`, `/context` | manage context usage |
 | `/permissions`, `/tools` | inspect available tools and policy |
@@ -310,8 +281,7 @@ zero verify           detect and run local verification checks
 zero changes          inspect and commit local git changes
 zero usage            token usage and estimated cost
 zero cron             scheduled agent jobs
-zero update --check   check for newer releases
-zero upgrade          download, verify, and install the latest release
+zero update           check for newer releases
 ```
 
 ## Extending Zero
@@ -352,7 +322,7 @@ manifest.
 | Control | Effect |
 |---|---|
 | `NO_COLOR=<anything>` | disables color output |
-| `ZERO_THEME=<name>` | selects the startup theme (`auto`, `dark`, `light`, or a color theme like `dracula`, `nord`, `gruvbox`, `tokyo-night`, `catppuccin`, `one-dark`, `solarized-dark`, `rose-pine`, `everforest`, `neon`, `solarized-light`, `dune`) |
+| `ZERO_THEME=<name>` | selects the startup theme (`auto`, `dark`, `light`, or a color theme like `dracula`, `nord`, `gruvbox`, `tokyo-night`, `catppuccin`, `one-dark`, `solarized-dark`, `rose-pine`, `everforest`, `solarized-light`) |
 | `--theme <name>` | selects the TUI theme from the CLI (same names) |
 | `/theme` | opens the theme picker inside the TUI (live preview; `/theme <name>` switches directly) |
 | `ZERO_NO_FADE=1` | disables streaming fade animation |
@@ -369,35 +339,7 @@ go run ./cmd/zero-release smoke
 go run ./cmd/zero-perf-bench
 ```
 
-Experimental: `ZERO_OPENAI_TURN_SESSION=1` enables the optimized OpenAI turn
-session (background connection prewarm + request-prefix telemetry) for headless
-`zero exec` runs against official OpenAI profiles. Off by default; `0`/`false`
-disable. A/B-benchmark it by running the same `zero-perf-bench` suite with the
-variable unset and set.
-
-### Code Quality and Security Checks
-
-Before committing any changes, ensure all Go code quality and security checks pass. The `make` targets below pin each tool to this module's Go version, so they load correctly even when your default `go` toolchain is older — running the plain `go run ...@version` form yourself can select the tool module's own (older) toolchain and fail to load this module instead.
-
-1. **Formatting**: Run `make fmt` (or `go fmt ./...`).
-2. **Vetting**: Run `make vet` (or `go vet ./...`).
-3. **Linting**: Run `make lint-static`.
-4. **Vulnerability Scan**: Run `make vulncheck`.
-
-Use the repository-managed targets rather than globally installed binaries:
-the targets apply the module's required Go toolchain as well as the reviewed,
-pinned tool versions.
-
-The installed binaries land in `$GOBIN` when it is set, otherwise in
-`$GOPATH/bin` (default `~/go/bin`). That directory must be on your `PATH` to
-run them directly. If it isn't, add it:
-
-```bash
-gobin="$(go env GOBIN)"; [ -z "$gobin" ] && gobin="$(go env GOPATH)/bin"
-case ":$PATH:" in *":$gobin:"*) ;; *) export PATH="$PATH:$gobin" ;; esac
-```
-
-### Cross-Compile Examples
+Cross-compile examples:
 
 ```bash
 go run ./cmd/zero-release build --goos linux --goarch amd64
@@ -408,7 +350,6 @@ go run ./cmd/zero-release build --goos windows --goarch amd64 --output dist/zero
 
 - [Install](docs/INSTALL.md)
 - [Update flow](docs/UPDATE.md)
-- [Themes](docs/THEMES.md)
 - [Stream-JSON protocol](docs/STREAM_JSON_PROTOCOL.md)
 - [Specialists](docs/SPECIALISTS.md)
 - [GitHub Action](docs/GITHUB_ACTION.md)
@@ -416,26 +357,6 @@ go run ./cmd/zero-release build --goos windows --goarch amd64 --output dist/zero
 - [Performance](docs/PERFORMANCE.md)
 - [Agent evals](docs/AGENT_EVALS.md)
 
-## Community
-
-Real-time chat happens on the [Discord server](https://discord.gg/CaQDS6wdFn).
-
-Questions, setup help, ideas, and sharing all live in
-[GitHub Discussions](https://github.com/Gitlawb/zero/discussions):
-
-| Category | Use it for |
-|---|---|
-| [Q&A](https://github.com/Gitlawb/zero/discussions/categories/q-a) | Setup help, provider/model configuration, "how do I" questions |
-| [Ideas](https://github.com/Gitlawb/zero/discussions/categories/ideas) | Feature proposals and design discussion before any PR |
-| [Show and tell](https://github.com/Gitlawb/zero/discussions/categories/show-and-tell) | Your skills, plugins, MCP setups, themes, and workflows |
-| [Announcements](https://github.com/Gitlawb/zero/discussions/categories/announcements) | Releases and project news from the maintainers |
-
-For a good Q&A answer fast, include `zero --version`, your OS and install
-method, the provider/model in use, and `zero doctor` output. See
-[SUPPORT.md](SUPPORT.md). Bugs belong in
-[issues](https://github.com/Gitlawb/zero/issues/new/choose); security reports
-follow [SECURITY.md](SECURITY.md), never a public thread.
-
 ## Contributing
 
 Contributions are welcome. Read [CONTRIBUTING.md](CONTRIBUTING.md), run the

README_ZH.md

@@ -6,9 +6,8 @@
 
 <p align="center">
   <a href="LICENSE"><img alt="license" src="https://img.shields.io/badge/license-MIT-blue"></a>
-  <img alt="Go 1.26.5+" src="https://img.shields.io/badge/Go-1.26.5+-00ADD8?logo=go&logoColor=white">
+  <img alt="Go 1.25+" src="https://img.shields.io/badge/Go-1.25+-00ADD8?logo=go&logoColor=white">
   <img alt="25+ providers" src="https://img.shields.io/badge/providers-25+-34E2EA">
-  <a href="https://discord.gg/CaQDS6wdFn"><img alt="Discord" src="https://img.shields.io/badge/Discord-join-5865F2?logo=discord&logoColor=white"></a>
   <br>
   <a href="README.md">English</a> | <strong>中文</strong>
 </p>
@@ -39,7 +38,25 @@ npm install -g @gitlawb/zero
 zero
 ```
 
-npm 包是一个小型包装器，其平台构建（Linux 和 macOS 的 x64/arm64、Windows 的 x64，包含浏览器/终端控制辅助工具）作为可选依赖直接从 npm registry 安装——没有安装脚本，也不会从 npm 之外下载任何内容。Bun、pnpm 和 yarn 的行为完全一致，无需任何信任或批准步骤。跳过可选依赖的安装（`--omit=optional`）也能工作：只要二进制文件缺失，包装器就会从对应的 GitHub Release 下载。Windows on ARM 通过模拟运行 x64 构建。详见 [docs/NPM_PACKAGING.md](docs/NPM_PACKAGING.md)。
+npm 包安装一个小型包装器以及与你平台匹配的 Zero 二进制文件（从 GitHub Releases 获取）。支持 Linux、macOS 和 Windows 的 x64 和 arm64。
+
+### Bun
+
+Bun 默认不运行依赖的生命周期脚本，因此获取 Zero 二进制文件的 `postinstall` 会被跳过，首次运行会失败并显示 `No native binary found next to the npm wrapper`。
+
+最简单的解决方法是安装后信任该包，这会运行被阻止的 postinstall。项目安装和全局安装均适用：
+
+```bash
+# 项目安装
+bun add @gitlawb/zero
+bun pm trust @gitlawb/zero
+
+# 全局安装
+bun add -g @gitlawb/zero
+bun pm -g trust @gitlawb/zero
+```
+
+其他方式：在 `bun add` 之前将 `"trustedDependencies": ["@gitlawb/zero"]` 添加到项目的 package.json 以提前允许 postinstall；或在不支持 `bun pm trust` 的旧版 Bun 上手动运行安装程序（`node node_modules/@gitlawb/zero/scripts/postinstall.mjs`）。
 
 ### 安装脚本
 
@@ -57,7 +74,7 @@ irm https://raw.githubusercontent.com/Gitlawb/zero/main/scripts/install.ps1 | ie
 
 ### 从源码构建
 
-源码构建需要 Go 1.26.5+。
+源码构建需要 Go 1.25+。
 
 ```bash
 git clone https://github.com/Gitlawb/zero.git
@@ -142,7 +159,6 @@ zero
 | `/spec`、`/plan` | 在构建之前起草和审查计划 |
 | `/image` | 为视觉模型附加图片 |
 | `/resume`、`/rewind` | 继续或回滚本地会话 |
-| `/new` | 原地开始一个新会话（之前的会话仍保留在磁盘上） |
 | `/compact`、`/context` | 管理上下文使用 |
 | `/permissions`、`/tools` | 检查可用工具和策略 |
 | `/add-dir` | 为此会话授予额外的写入目录 |
@@ -228,8 +244,7 @@ zero verify           检测和运行本地验证检查
 zero changes          检查和提交本地 Git 变更
 zero usage            Token 使用量和估算成本
 zero cron             定时智能体任务
-zero update --check   检查更新版本
-zero upgrade          下载、验证并安装最新版本
+zero update           检查更新版本
 ```
 
 ## 外观和无障碍
@@ -237,7 +252,7 @@ zero upgrade          下载、验证并安装最新版本
 | 控制 | 效果 |
 |---|---|
 | `NO_COLOR=<任意值>` | 禁用颜色输出 |
-| `ZERO_THEME=<名称>` | 选择启动主题（`auto`、`dark`、`light`，或颜色主题如 `dracula`、`nord`、`gruvbox`、`tokyo-night`、`catppuccin`、`one-dark`、`solarized-dark`、`rose-pine`、`everforest`、`neon`、`solarized-light`、`dune`） |
+| `ZERO_THEME=<名称>` | 选择启动主题（`auto`、`dark`、`light`，或颜色主题如 `dracula`、`nord`、`gruvbox`、`tokyo-night`、`catppuccin`、`one-dark`、`solarized-dark`、`rose-pine`、`everforest`、`solarized-light`） |
 | `--theme <名称>` | 从 CLI 选择 TUI 主题（相同名称） |
 | `/theme` | 在 TUI 中打开主题选择器（实时预览；`/theme <名称>` 直接切换） |
 | `ZERO_NO_FADE=1` | 禁用流式淡入动画 |
@@ -264,7 +279,6 @@ go run ./cmd/zero-release build --goos windows --goarch amd64 --output dist/zero
 
 - [安装](docs/INSTALL.md)
 - [更新流程](docs/UPDATE.md)
-- [主题](docs/THEMES.md)
 - [Stream-JSON 协议](docs/STREAM_JSON_PROTOCOL.md)
 - [专家](docs/SPECIALISTS.md)
 - [GitHub Action](docs/GITHUB_ACTION.md)
@@ -272,20 +286,6 @@ go run ./cmd/zero-release build --goos windows --goarch amd64 --output dist/zero
 - [性能](docs/PERFORMANCE.md)
 - [智能体评估](docs/AGENT_EVALS.md)
 
-## 社区
-
-实时交流在 [Discord 服务器](https://discord.gg/CaQDS6wdFn) 进行。
-
-提问、安装帮助、想法和分享都在
-[GitHub Discussions](https://github.com/Gitlawb/zero/discussions)：
-
-| 分类 | 用途 |
-|---|---|
-| [Q&A](https://github.com/Gitlawb/zero/discussions/categories/q-a) | 安装帮助、提供商/模型配置、"如何做"类问题 |
-| [Ideas](https://github.com/Gitlawb/zero/discussions/categories/ideas) | 功能提议和 PR 之前的设计讨论 |
-| [Show and tell](https://github.com/Gitlawb/zero/discussions/categories/show-and-tell) | 你的技能、插件、MCP 配置、主题和工作流 |
-| [Announcements](https://github.com/Gitlawb/zero/discussions/categories/announcements) | 维护者发布的版本和项目动态 |
-
 ## 贡献
 
 欢迎贡献。阅读 [CONTRIBUTING.md](CONTRIBUTING.md)，运行相关测试，然后提交一个聚焦的拉取请求。

SUPPORT.md

@@ -1,36 +0,0 @@
-# Getting help with Zero
-
-## Questions, setup help, configuration
-
-Ask in [GitHub Discussions](https://github.com/Gitlawb/zero/discussions). Use
-[Q&A](https://github.com/Gitlawb/zero/discussions/categories/q-a) so answers can
-be marked and found by the next person.
-
-The fastest way to a useful answer is to include:
-
-- `zero --version` output
-- your OS (Windows, macOS, Linux) and how you installed Zero (npm, install
-  script, source)
-- the provider and model you are using (`zero providers list`)
-- `zero doctor` output for setup problems
-- the exact command you ran and the exact error text
-
-## Bugs
-
-If Zero does something wrong (crashes, wrong output, broken behavior), file a
-[bug report](https://github.com/Gitlawb/zero/issues/new/choose). If you are not
-sure whether it is a bug or a setup problem, start in Discussions and we will
-promote it to an issue if it is one.
-
-## Feature ideas
-
-Start a thread in
-[Ideas](https://github.com/Gitlawb/zero/discussions/categories/ideas). Zero is
-in controlled development (see CONTRIBUTING.md), so discussing an idea first is
-also the path to getting a PR accepted.
-
-## Security issues
-
-Never file a public issue or discussion. Report privately via
-[GitHub Security Advisories](https://github.com/Gitlawb/zero/security/advisories/new)
-(see SECURITY.md).

bin/zero.js

@@ -2,7 +2,6 @@
 
 import { spawnSync } from 'node:child_process';
 import { existsSync } from 'node:fs';
-import { createRequire } from 'node:module';
 import { dirname, join } from 'node:path';
 import { fileURLToPath } from 'node:url';
 
@@ -40,11 +39,6 @@ function resolveHelper(packageRoot, name) {
   return null;
 }
 
-// Legacy path: helpers installed as npm dependencies of the wrapper. Current
-// installs vendor the helpers inside the platform package instead (a helpers/
-// directory next to the binary, found by the Go side on its own), so this
-// manifest is usually empty — it only matters for a downloaded-binary install
-// whose wrapper still carries helper dependencies.
 function localControlHelperManifest(packageRoot) {
   const helpers = {};
   for (const name of ['agent-browser', 'tuistory']) {
@@ -55,229 +49,38 @@ function localControlHelperManifest(packageRoot) {
   return JSON.stringify({ version: 1, helpers });
 }
 
-// Mirrors internal/cli/observability.go parseDoctorArgs + writeDoctorHelp so the
-// wrapper's missing-binary doctor fallback matches the Go doctor CLI surface.
-const DOCTOR_HELP = `Usage:
-  zero doctor [flags]
-
-Runs Go backend health checks for config and provider setup.
-
-Flags:
-      --json            Print JSON report
-      --connectivity    Include provider endpoint connectivity probe when available
-  -h, --help            Show this help
-`;
-
-const EXIT_USAGE = 2;
-
-function installedByBun() {
-  if (process.env.ZERO_WRAPPER_SIMULATE_BUN === '1') {
-    return true;
-  }
-  return process.execPath.includes('bun') || !!process.versions?.bun;
-}
-
-function bunRecoveryParagraph() {
-  return (
-    'You installed with Bun, which does not run dependency lifecycle scripts\n' +
-    'by default. Trust the package to run the blocked postinstall:\n' +
-    '  bun pm trust @gitlawb/zero       (project install)\n' +
-    '  bun pm -g trust @gitlawb/zero    (global install)\n' +
-    'On Bun versions without `bun pm trust`, add\n' +
-    '  "trustedDependencies": ["@gitlawb/zero"]\n' +
-    'to your project package.json and reinstall.\n' +
-    '\n'
-  );
-}
-
-function buildFromSourceParagraph() {
-  return (
-    'If this platform has no prebuilt binary, build from source:\n' +
-    'https://github.com/Gitlawb/zero (go run ./cmd/zero, requires Go 1.26+).'
-  );
-}
-
-function missingBinaryContextParagraph() {
-  return (
-    'Normally npm installs it as an optional dependency of @gitlawb/zero\n' +
-    `(@gitlawb/zero-${process.platform}-${process.arch}), and the wrapper can\n` +
-    'also download it from the GitHub Release when it is missing.'
-  );
-}
-
-function missingNativeRecoveryParagraphs(postinstallScript) {
-  const ranByBun = installedByBun();
-  return (
-    missingBinaryContextParagraph() +
-    '\n' +
-    '\n' +
-    'Things to try:\n' +
-    '  - reinstall without omitting optional dependencies:\n' +
-    '      npm install -g @gitlawb/zero\n' +
-    '  - run the downloader manually (needs write access to the package dir):\n' +
-    `      node "${postinstallScript}"\n` +
-    '\n' +
-    (ranByBun ? bunRecoveryParagraph() : '') +
-    buildFromSourceParagraph()
-  );
-}
-
-function formatGenericMissingNativeBinaryMessage(postinstallScript) {
-  return (
-    '[zero] No native binary is available for this install.\n' +
-    missingNativeRecoveryParagraphs(postinstallScript)
-  );
-}
-
-function parseDoctorArgs(args) {
-  let json = false;
-  for (const arg of args) {
-    switch (arg) {
-      case '-h':
-      case '--help':
-      case 'help':
-        return { kind: 'help' };
-      case '--json':
-        json = true;
-        break;
-      case '--connectivity':
-        break;
-      default:
-        return { kind: 'error', message: `unknown doctor flag ${JSON.stringify(arg)}` };
-    }
-  }
-  return { kind: 'run', json };
-}
-
-function missingNativeDoctorJSONReport(postinstallScript) {
-  return {
-    generatedAt: new Date().toISOString().replace(/\.\d{3}Z$/, 'Z'),
-    ok: false,
-    checks: [
-      {
-        id: 'runtime.go',
-        label: 'Go runtime',
-        status: 'fail',
-        message: 'Native zero binary is missing next to the npm wrapper.',
-        details: {
-          remedy: `node "${postinstallScript}"`,
-        },
-      },
-    ],
-  };
-}
-
-function missingNativeDoctorTextReport(postinstallScript) {
-  return (
-    'Zero doctor report (' +
-    new Date().toISOString() +
-    ')\n' +
-    'Overall: fail\n' +
-    '[fail] runtime.go - Native zero binary is missing next to the npm wrapper.\n' +
-    '  remedy: node "' +
-    postinstallScript +
-    '"\n' +
-    '\n' +
-    missingNativeRecoveryParagraphs(postinstallScript)
-  );
-}
-
-// The platform payload is a version of @gitlawb/zero itself, installed under
-// an npm: alias (see docs/NPM_PACKAGING.md). The alias name is derived from
-// process.platform/process.arch, so an unsupported platform simply fails to
-// resolve and we fall through to the downloader.
-function platformPackageBinary() {
-  const aliasedName = `@gitlawb/zero-${process.platform}-${process.arch}`;
-  let manifestPath;
-  try {
-    manifestPath = createRequire(import.meta.url).resolve(`${aliasedName}/package.json`);
-  } catch {
-    return null;
-  }
-  const candidate = join(dirname(manifestPath), zeroBinaryName());
-  return existsSync(candidate) ? candidate : null;
-}
-
-// Mirrors the release matrix (internal/release/release.go): the platforms a
-// GitHub Release asset exists for. android maps to the linux asset; there is
-// no windows-arm64 asset (Windows on ARM runs the x64 build under emulation).
-function releaseAssetAvailable() {
-  const platform = { linux: 'linux', android: 'linux', darwin: 'macos', win32: 'windows' }[
-    process.platform
-  ];
-  const arch = process.arch === 'x64' || process.arch === 'arm64' ? process.arch : null;
-  if (!platform || !arch) return false;
-  return !(platform === 'windows' && arch === 'arm64');
-}
-
-// Fallback for installs without a platform package (--omit=optional, package
-// managers that skip optionalDependencies): fetch the binary next to the
-// wrapper with the same checksum-verified downloader that used to run as
-// postinstall. Deliberately NOT cached on failure — a transient network error
-// self-heals on the next run — so this retries on every invocation until a
-// binary is in place. Platforms with no release asset skip the attempt.
-function downloadMissingBinary(packageRoot) {
-  const downloadScript = join(packageRoot, 'scripts', 'postinstall.mjs');
-  if (!existsSync(downloadScript) || !releaseAssetAvailable()) return;
-  console.error(
-    '[zero] no platform package installed — fetching the native binary from the GitHub Release (retried on each run until it succeeds).',
-  );
-  spawnSync(process.execPath, [downloadScript], {
-    stdio: ['ignore', 'inherit', 'inherit'],
-  });
-}
-
 const packageRoot = dirname(dirname(fileURLToPath(import.meta.url)));
+const nativePath = join(packageRoot, zeroBinaryName());
+const localControlHelpers = localControlHelperManifest(packageRoot);
 
-function resolveNativeBinary() {
-  const fromPlatformPackage = platformPackageBinary();
-  if (fromPlatformPackage) return fromPlatformPackage;
-  const downloaded = join(packageRoot, zeroBinaryName());
-  if (existsSync(downloaded)) return downloaded;
-  downloadMissingBinary(packageRoot);
-  return existsSync(downloaded) ? downloaded : null;
-}
-
-const nativePath = resolveNativeBinary();
-
-if (!nativePath) {
+if (!existsSync(nativePath)) {
   const postinstallScript = join(packageRoot, 'scripts', 'postinstall.mjs');
-
-  // `zero doctor` is the diagnostic command: when the native binary is missing
-  // it's the one invocation that must NOT bail with the generic wrapper error,
-  // because that's exactly the blind alley issue #405 reports. Instead, surface
-  // a doctor-shaped FAIL line for the runtime so the user's diagnostic finds the
-  // real cause. We branch on a literal 'doctor' subcommand only (matching `zero
-  // doctor` and `zero doctor --connectivity`), preserving the existing bail for
-  // every other invocation (exec, providers list, TUI, --version, etc.).
-  const argv = process.argv.slice(2);
-  const isDoctor = argv.length > 0 && argv[0] === 'doctor';
-  if (isDoctor) {
-    const parsed = parseDoctorArgs(argv.slice(1));
-    if (parsed.kind === 'help') {
-      process.stdout.write(DOCTOR_HELP);
-      process.exit(0);
-    }
-    if (parsed.kind === 'error') {
-      process.stderr.write(`[zero] ${parsed.message}\n`);
-      process.exit(EXIT_USAGE);
-    }
-
-    if (parsed.json) {
-      process.stdout.write(JSON.stringify(missingNativeDoctorJSONReport(postinstallScript), null, 2) + '\n');
-      process.exit(1);
-    }
-
-    process.stdout.write(missingNativeDoctorTextReport(postinstallScript) + '\n');
-    process.exit(1);
-  }
-
-  console.error(formatGenericMissingNativeBinaryMessage(postinstallScript));
+  const ranByBun = process.execPath.includes('bun') || !!process.versions?.bun;
+  console.error(
+    '[zero] No native binary found next to the npm wrapper.\n' +
+      'The platform binary is fetched at install time by a postinstall script,\n' +
+      'which did not run (or was skipped) for this install.\n' +
+      '\n' +
+      'Fix it now by running the installer manually:\n' +
+      `  node "${postinstallScript}"\n` +
+      '\n' +
+      (ranByBun
+        ? 'You installed with Bun, which does not run dependency lifecycle scripts\n' +
+          'by default. Trust the package to run the blocked postinstall:\n' +
+          '  bun pm trust @gitlawb/zero       (project install)\n' +
+          '  bun pm -g trust @gitlawb/zero    (global install)\n' +
+          'On Bun versions without `bun pm trust`, add\n' +
+          '  "trustedDependencies": ["@gitlawb/zero"]\n' +
+          'to your project package.json and reinstall.\n' +
+          '\n'
+        : '') +
+      'If that fails, build from source: https://github.com/Gitlawb/zero\n' +
+      '(go run ./cmd/zero, requires Go 1.25+).',
+  );
   process.exit(1);
 }
 
 const env = { ...process.env };
-const localControlHelpers = localControlHelperManifest(packageRoot);
 if (localControlHelpers) {
   env.ZERO_LOCAL_CONTROL_HELPERS = localControlHelpers;
 } else {

cmd/zero-perf-bench/main.go

@@ -39,9 +39,6 @@ func run(args []string, getenv func(string) string, stdout io.Writer, stderr io.
 	if len(args) > 0 && args[0] == "tasks" {
 		return runTasksCommand(args[1:], getenv, stdout, stderr)
 	}
-	if len(args) > 0 && args[0] == "turn" {
-		return runTurnCommand(args[1:], getenv, stdout, stderr)
-	}
 	options, err := parseArgs(args, getenv)
 	if err != nil {
 		_, _ = fmt.Fprintln(stderr, err.Error())
@@ -209,7 +206,6 @@ func helpText() string {
 	return strings.Join([]string{
 		"Usage: zero-perf-bench [options]",
 		"       zero-perf-bench tasks [options]   (Terminal-Bench-style task harness; see `tasks --help`)",
-		"       zero-perf-bench turn  [options]   (per-turn tracing benchmark; see `turn --help`)",
 		"",
 		"Options:",
 		"  --iterations <n>             Measured samples to collect (default: 5)",

cmd/zero-perf-bench/main_test.go

@@ -4,8 +4,6 @@ import (
 	"bytes"
 	"strings"
 	"testing"
-
-	"github.com/Gitlawb/zero/internal/perfbench"
 )
 
 func TestParseArgsUsesCliAndEnvOverrides(t *testing.T) {
@@ -94,18 +92,3 @@ func TestHelpText(t *testing.T) {
 func emptyEnv(string) string {
 	return ""
 }
-
-func TestTurnExitCodeFailsWhenEveryTaskErrored(t *testing.T) {
-	var stderr bytes.Buffer
-	allErrored := perfbench.TurnBenchResult{TasksAttempted: 3, TasksErrored: 3}
-	if code := turnExitCode(allErrored, &stderr); code != 1 {
-		t.Fatalf("exit=%d, want 1 when every task errored", code)
-	}
-	if !strings.Contains(stderr.String(), "no accepted benchmark sample") {
-		t.Fatalf("stderr missing explanation: %q", stderr.String())
-	}
-	partial := perfbench.TurnBenchResult{TasksAttempted: 3, TasksErrored: 2}
-	if code := turnExitCode(partial, &stderr); code != 0 {
-		t.Fatalf("exit=%d, want 0 for partial errors (summary surfaces them)", code)
-	}
-}

cmd/zero-perf-bench/turn.go

@@ -1,279 +0,0 @@
-package main
-
-import (
-	"bytes"
-	"context"
-	"fmt"
-	"io"
-	"os"
-	"path/filepath"
-	"strings"
-
-	"github.com/Gitlawb/zero/internal/execprofile"
-	"github.com/Gitlawb/zero/internal/perfbench"
-)
-
-// turnOptions configures the `zero-perf-bench turn` subcommand: the per-turn
-// benchmark harness that runs ZERO headlessly with --trace, parses each turn's
-// NDJSON trace, and records per-span latency plus the top controllable latency
-// sources — the Phase 0 baseline's "do not proceed until" criterion.
-type turnOptions struct {
-	SuitePath   string
-	Model       string
-	Mode        string
-	ExecProfile string
-	SelfCorrect bool
-	Binary      string
-	Iterations  int
-	Version     string
-	Commit      string
-	Output      string
-	JSON        bool
-	DryRun      bool
-	Help        bool
-}
-
-func runTurnCommand(args []string, getenv func(string) string, stdout io.Writer, stderr io.Writer) int {
-	options, err := parseTurnArgs(args, getenv)
-	if err != nil {
-		_, _ = fmt.Fprintln(stderr, err.Error())
-		return 2
-	}
-	if options.Help {
-		_, _ = fmt.Fprint(stdout, turnHelpText())
-		return 0
-	}
-
-	set, err := perfbench.LoadTaskSet(options.SuitePath)
-	if err != nil {
-		_, _ = fmt.Fprintln(stderr, "[zero] Turn benchmark failed: "+err.Error())
-		return 1
-	}
-
-	// The dry-run path records a zero-iteration run without a binary, so the
-	// manifest loads and the report path is exercised in CI without a model.
-	if options.DryRun {
-		_, _ = fmt.Fprintln(stdout, "[zero] turn benchmark: dry run (no agent invoked)")
-		return 0
-	}
-
-	binary, err := perfbench.ResolveBinary(options.Binary)
-	if err != nil {
-		_, _ = fmt.Fprintln(stderr, "[zero] Turn benchmark failed: "+err.Error())
-		return 2
-	}
-
-	result, err := perfbench.RunTurnBench(context.Background(), set, perfbench.TurnBenchConfig{
-		Model:       options.Model,
-		Mode:        options.Mode,
-		ExecProfile: options.ExecProfile,
-		SelfCorrect: options.SelfCorrect,
-		Version:     options.Version,
-		Commit:      options.Commit,
-		Iterations:  options.Iterations,
-		Runner:      perfbench.NewTurnExecRunner(binary),
-	})
-	if err != nil {
-		_, _ = fmt.Fprintln(stderr, "[zero] Turn benchmark failed: "+err.Error())
-		return 1
-	}
-
-	if options.Output != "" {
-		if err := writeTurnReport(options.Output, result); err != nil {
-			_, _ = fmt.Fprintln(stderr, "[zero] Turn benchmark failed: "+err.Error())
-			return 1
-		}
-	}
-	if options.JSON {
-		if err := perfbench.WriteTurnBenchJSON(stdout, result); err != nil {
-			_, _ = fmt.Fprintln(stderr, "[zero] Turn benchmark failed: "+err.Error())
-			return 1
-		}
-		return turnExitCode(result, stderr)
-	}
-	_, _ = fmt.Fprintln(stdout, perfbench.FormatTurnBenchSummary(result))
-	return turnExitCode(result, stderr)
-}
-
-// turnExitCode fails the command when every attempted task errored (no
-// iteration produced an accepted benchmark sample): such a report measures
-// nothing, and exiting 0 would let a broken configuration (missing binary, bad
-// path, failing harness step) pass as a clean baseline. Partial errors keep
-// exit 0 — the summary surfaces them loudly and the surviving samples are
-// still valid measurements.
-func turnExitCode(result perfbench.TurnBenchResult, stderr io.Writer) int {
-	if result.TasksAttempted > 0 && result.TasksErrored == result.TasksAttempted {
-		_, _ = fmt.Fprintln(stderr, "[zero] Turn benchmark failed: every task errored with no accepted benchmark sample (see warnings); the report contains no valid measurements")
-		return 1
-	}
-	return 0
-}
-
-func parseTurnArgs(args []string, getenv func(string) string) (turnOptions, error) {
-	options := turnOptions{
-		Iterations: 1,
-		Version:    strings.TrimSpace(getenv("ZERO_BENCH_VERSION")),
-		Commit:     strings.TrimSpace(getenv("ZERO_BENCH_COMMIT")),
-	}
-	for index := 0; index < len(args); index++ {
-		arg := args[index]
-		flag, inlineValue := splitFlagValue(arg)
-		switch flag {
-		case "--suite":
-			value, next, err := readOptionValue(args, inlineValue, index, flag)
-			if err != nil {
-				return options, err
-			}
-			options.SuitePath = value
-			index = next
-		case "--model":
-			value, next, err := readOptionValue(args, inlineValue, index, flag)
-			if err != nil {
-				return options, err
-			}
-			options.Model = value
-			index = next
-		case "--mode":
-			value, next, err := readOptionValue(args, inlineValue, index, flag)
-			if err != nil {
-				return options, err
-			}
-			options.Mode = value
-			index = next
-		case "--exec-profile":
-			value, next, err := readOptionValue(args, inlineValue, index, flag)
-			if err != nil {
-				return options, err
-			}
-			options.ExecProfile = value
-			index = next
-		case "--binary":
-			value, next, err := readOptionValue(args, inlineValue, index, flag)
-			if err != nil {
-				return options, err
-			}
-			options.Binary = value
-			index = next
-		case "--iterations":
-			value, next, err := readOptionValue(args, inlineValue, index, flag)
-			if err != nil {
-				return options, err
-			}
-			parsed, err := parsePositiveInteger(flag, value)
-			if err != nil {
-				return options, err
-			}
-			options.Iterations = parsed
-			index = next
-		case "--version":
-			value, next, err := readOptionValue(args, inlineValue, index, flag)
-			if err != nil {
-				return options, err
-			}
-			options.Version = value
-			index = next
-		case "--commit":
-			value, next, err := readOptionValue(args, inlineValue, index, flag)
-			if err != nil {
-				return options, err
-			}
-			options.Commit = value
-			index = next
-		case "--output":
-			value, next, err := readOptionValue(args, inlineValue, index, flag)
-			if err != nil {
-				return options, err
-			}
-			options.Output = value
-			index = next
-		case "--self-correct":
-			if strings.Contains(arg, "=") {
-				return options, fmt.Errorf("%s does not accept a value", flag)
-			}
-			options.SelfCorrect = true
-		case "--json":
-			if strings.Contains(arg, "=") {
-				return options, fmt.Errorf("%s does not accept a value", flag)
-			}
-			options.JSON = true
-		case "--dry-run":
-			if strings.Contains(arg, "=") {
-				return options, fmt.Errorf("%s does not accept a value", flag)
-			}
-			options.DryRun = true
-		case "-h", "--help":
-			if strings.Contains(arg, "=") {
-				return options, fmt.Errorf("%s does not accept a value", flag)
-			}
-			options.Help = true
-		default:
-			return options, fmt.Errorf("unknown option: %s", arg)
-		}
-	}
-	if options.Help {
-		return options, nil
-	}
-	if strings.TrimSpace(options.SuitePath) == "" {
-		return options, fmt.Errorf("--suite is required")
-	}
-	if strings.TrimSpace(options.Model) == "" && !options.DryRun {
-		return options, fmt.Errorf("--model is required (or pass --dry-run)")
-	}
-	// Validate the profile here, before anything spawns: an unknown name would
-	// otherwise make every child exit with a usage error in milliseconds, and
-	// those near-zero walls would be recorded as valid latency samples in an
-	// exit-0 report — exactly the misread-as-improvement trap the harness
-	// closed for spawn failures. Normalizing to the catalog name also keeps
-	// the stamped execProfile canonical (Lookup is case-insensitive), so two
-	// captures of the same posture always compare equal.
-	if raw := strings.TrimSpace(options.ExecProfile); raw != "" {
-		profile, ok := execprofile.Lookup(raw)
-		if !ok {
-			return options, fmt.Errorf("unknown execution profile %q for --exec-profile. Valid profiles: %s", raw, strings.Join(execprofile.Names(), ", "))
-		}
-		options.ExecProfile = profile.Name
-	}
-	return options, nil
-}
-
-func writeTurnReport(path string, result perfbench.TurnBenchResult) error {
-	if dir := filepath.Dir(path); dir != "" {
-		if err := os.MkdirAll(dir, 0o755); err != nil {
-			return err
-		}
-	}
-	var buffer bytes.Buffer
-	if err := perfbench.WriteTurnBenchJSON(&buffer, result); err != nil {
-		return err
-	}
-	return os.WriteFile(path, buffer.Bytes(), 0o644)
-}
-
-func turnHelpText() string {
-	return strings.Join([]string{
-		"Usage: zero-perf-bench turn [options]",
-		"",
-		"Runs ZERO headlessly with --trace against a per-turn benchmark task set and",
-		"records per-span latency plus the top controllable latency sources (the",
-		"Phase 0 baseline's \"do not proceed until\" criterion). Each task is a fresh",
-		"`zero exec` process, so iterations are cold-start samples; a warm path needs",
-		"an in-process runner (future work).",
-		"",
-		"Options:",
-		"  --suite <path>      Task set JSON file (required)",
-		"  --model <model>     Model to run (required unless --dry-run)",
-		"  --mode <name>       Exec mode preset to apply",
-		"  --exec-profile <name>",
-		"                      Execution profile for every task (fast|balanced|thorough);",
-		"                      forwarded to zero exec and stamped into the result",
-		"  --self-correct      Enable the post-edit verify-and-correct loop",
-		"  --binary <path>     Path to the `zero` binary (default: zero on PATH / repo root)",
-		"  --iterations <n>    Times to run each task (default: 1)",
-		"  --version <v>       Record the ZERO version (default: $ZERO_BENCH_VERSION)",
-		"  --commit <sha>      Record the ZERO commit (default: $ZERO_BENCH_COMMIT)",
-		"  --output <path>     Write the JSON result to path",
-		"  --json              Print only the JSON result",
-		"  --dry-run           Load the manifest and exit without invoking the agent",
-		"  -h, --help          Show this help",
-	}, "\n") + "\n"
-}

cmd/zero-perf-bench/turn_test.go

@@ -1,65 +0,0 @@
-package main
-
-import (
-	"strings"
-	"testing"
-)
-
-func TestParseTurnArgsExecProfile(t *testing.T) {
-	getenv := func(string) string { return "" }
-	options, err := parseTurnArgs([]string{
-		"--suite", "suite.json",
-		"--model", "m",
-		"--exec-profile", "fast",
-	}, getenv)
-	if err != nil {
-		t.Fatalf("parseTurnArgs: %v", err)
-	}
-	if options.ExecProfile != "fast" {
-		t.Fatalf("ExecProfile = %q, want fast", options.ExecProfile)
-	}
-
-	// Inline form.
-	options, err = parseTurnArgs([]string{
-		"--suite=suite.json",
-		"--model=m",
-		"--exec-profile=thorough",
-	}, getenv)
-	if err != nil {
-		t.Fatalf("parseTurnArgs inline: %v", err)
-	}
-	if options.ExecProfile != "thorough" {
-		t.Fatalf("ExecProfile = %q, want thorough", options.ExecProfile)
-	}
-}
-
-// The bench validates the profile BEFORE spawning anything: an unknown name
-// would otherwise make every child exit with a usage error in milliseconds
-// and record those near-zero walls as valid latency samples in an exit-0
-// report. Lookup normalization also keeps the stamped name canonical so two
-// captures of the same posture compare equal.
-func TestParseTurnArgsExecProfileValidatesAndNormalizes(t *testing.T) {
-	getenv := func(string) string { return "" }
-
-	options, err := parseTurnArgs([]string{
-		"--suite", "suite.json", "--model", "m", "--exec-profile", "FAST",
-	}, getenv)
-	if err != nil {
-		t.Fatalf("parseTurnArgs: %v", err)
-	}
-	if options.ExecProfile != "fast" {
-		t.Fatalf("ExecProfile = %q, want the canonical fast", options.ExecProfile)
-	}
-
-	_, err = parseTurnArgs([]string{
-		"--suite", "suite.json", "--model", "m", "--exec-profile", "blanced",
-	}, getenv)
-	if err == nil {
-		t.Fatal("an unknown profile must fail parse, before anything spawns")
-	}
-	for _, name := range []string{"balanced", "fast", "thorough"} {
-		if !strings.Contains(err.Error(), name) {
-			t.Fatalf("error must list %q, got %q", name, err.Error())
-		}
-	}
-}

docs/EXTENDING.md

@@ -1,341 +0,0 @@
-# Extending Zero
-
-Zero is an open-source terminal coding agent. Out of the box it does the obvious things — read, edit, run, search — but the design point of the project is that **every surface is configurable**. This document is the user-facing guide for that configuration.
-
-If you only want to *use* Zero, the [README](../README.md) is enough. This page is for the other three jobs:
-
-1. Tell the agent about *your* project (drop an `AGENTS.md` in your repo).
-2. Add new specialist sub-agents.
-3. Wire Zero into the rest of your toolchain (MCP, skills, hooks, plugins).
-
-## 1. Drop a project `AGENTS.md`
-
-When Zero starts in a directory, it looks for project-level instructions and injects them into the system prompt. The lookup walks from your current working directory **up to the nearest git root** and reads the first matching file at each level — general rules at the repo root, more specific rules in sub-trees. Files are labeled with their directory in the prompt (e.g. `## Project guidelines (services/api/AGENTS.md)`).
-
-Accepted file names, in priority order at each level:
-
-| Path | Notes |
-| --- | --- |
-| `./AGENTS.md` | The classic spot — committed to your repo, shared with the team. |
-| `./ZERO.md` | Brand-specific alias. Same format, lower priority. |
-| `./.zero/AGENTS.md` | Project-local, hidden, gitignored. Personal notes that stay out of git. |
-
-Matching is **case-insensitive** on the basename, so `AGENTS.md`, `Agents.md`, and `agents.md` resolve to the same file on Windows and macOS. The git-tracked filename in this repo is `AGENTS.md` — keep that on case-sensitive filesystems (Linux, the WSL filesystem, or a CI runner) to match what the loader looks for.
-
-Both files use the same format. YAML frontmatter is optional; the markdown body is loaded as instructions for the agent. Zero reads the file once at session start, so changes take effect on the next `zero` launch — not mid-session.
-
-```markdown
-# Project conventions for <your project>
-
-- Build with `make`, not `go build` directly.
-- Tests live next to the source file (`foo_test.go` next to `foo.go`).
-- Run `make lint` before opening a PR.
-- Never edit files under `third_party/` — those are vendored.
-```
-
-Tips:
-
-- Keep each file under ~8 KiB. Zero caps the **total** across all matched files at 32 KiB; everything past the cap is dropped.
-- Re-state rules in the imperative voice: "Run `make lint`", not "you should consider running the linter".
-- Don't put secrets, model IDs, or environment-specific paths in `AGENTS.md`. Use `config.json` for those.
-- In a monorepo, drop a narrower `AGENTS.md` in each sub-tree (e.g. `services/api/AGENTS.md`). Zero picks those up automatically when you launch from inside the sub-tree.
-- A YAML frontmatter block (`---\n...\n---`) at the top is preserved verbatim in the injected prompt but is not parsed for `globs:` or `alwaysApply:` scoping today — keep the body self-contained.
-
-### Personal guidelines, across every project
-
-For preferences that follow *you*, not a specific repo (tone, tooling habits, workflow), drop a `ZERO.md` in your user config directory: `~/.config/zero/ZERO.md` on Linux/macOS, `%AppData%\zero\ZERO.md` on Windows — the same directory as `config.json` and your personal specialists. Same format and 8 KiB cap as the project files above, and the same case-insensitive basename match.
-
-This file is injected as its own `## User guidelines` section, before the project's `AGENTS.md`/`ZERO.md`, and is labeled as personal preference in the prompt: project guidelines are the later, more specific instruction and take precedence over it when the two conflict.
-
-## 2. Custom specialists
-
-Specialists are Zero's sub-agents. Three scopes, in priority order:
-
-| Scope | Path | Shared? |
-| --- | --- | --- |
-| Built-in | compiled into Zero | yes — `worker`, `explorer`, `code-review` |
-| User | `~/.config/zero/specialists/*.md` | no — your machine only |
-| Project | `./.zero/specialists/*.md` | yes — the repo team |
-
-Project overrides user overrides built-in when names collide.
-
-A specialist is a markdown manifest with frontmatter and a system prompt:
-
-```markdown
----
-description: Reviews API changes for breaking-change risk and missing tests.
-tools: read-only,plan
----
-
-You review API changes. For every changed hunk in `internal/api/` or any file
-that ends in `_api.go`:
-
-1. Confirm the public signature is backward-compatible, or note the breaking
-   change explicitly with the migration path.
-2. Confirm a corresponding test exists in `internal/api/*_test.go` and that
-   the new behaviour is exercised.
-3. Flag any new exported symbol without a doc comment.
-
-Reply with one JSON object per finding: `{"file", "line", "severity", "message", "fix"}`.
-```
-
-CLI management (the prompt is passed inline via `--prompt`):
-
-```bash
-zero specialist list
-zero specialist show api-reviewer
-zero specialist create api-reviewer \
-    --project \
-    --description "Reviews API changes" \
-    --tools read-only,plan \
-    --prompt "$(cat api-reviewer.md)"
-zero specialist edit api-reviewer --project
-zero specialist delete api-reviewer --project
-zero specialist path                       # prints the resolved specialists directory
-```
-
-The full format spec (frontmatter fields, tool scopes, prompt conventions) is in [`docs/SPECIALISTS.md`](SPECIALISTS.md).
-
-> **Roadmap.** An in-UI specialist manager (create / edit / delete / preview) is on the backlog. Today you use the `zero specialist` CLI subcommands above.
-
-## 3. Skills
-
-Skills are markdown instruction packs the agent can pull in on demand. Each skill is a directory containing a `SKILL.md`. Standalone project skill directories are not supported in this version (shared project-wide skills must go in `AGENTS.md` or as a hook). However, project plugins (section 6) may bundle skills, which are merged into the active run.
-
-Discovery roots (earlier wins on name collisions):
-
-1. **Primary Zero dir** — `$ZERO_SKILLS_DIR` if set, else `$XDG_DATA_HOME/zero/skills`, else `~/.local/share/zero/skills/`
-2. **Shared multi-agent dir** — `~/.agents/skills/` when present (read-only discovery; never an install target)
-3. **Plugin skill roots** — skills bundled by active plugins (section 6)
-
-A missing directory is fine — Zero just omits it. Management commands (`zero skill add` / `zero skill remove`) always write to the primary Zero dir only; `list` / `info` search primary + `~/.agents/skills`. `skill info` reports lock source/hash when present and whether the on-disk `SKILL.md` has drifted from the recorded hash.
-
-```text
-~/.local/share/zero/skills/          # primary (install/remove target)
-  run-benchmarks/
-    SKILL.md
-  write-changelog/
-    SKILL.md
-
-~/.agents/skills/                    # optional shared multi-agent root
-  shared-review/
-    SKILL.md
-```
-
-`SKILL.md` format:
-
-```markdown
----
-description: Run the project's benchmark suite and summarize the deltas.
----
-
-# Run benchmarks
-
-1. `make bench` — captures the wall-clock and RSS before and after.
-2. `benchstat before.txt after.txt` — diffs the two.
-3. Report any regression > 5% with the function name and the previous value.
-```
-
-Only `name` and `description` are recognized in the frontmatter today. The `name` defaults to the directory name. Within a single skills root, duplicate names are resolved by lexicographic directory order. Across roots, Zero loads the primary directory first, then `~/.agents/skills`, then plugin skill roots; earlier roots win name collisions. Plugin-declared skills (section 6) are merged into the active agent run at plugin activation time, so bundled skills appear in the available skills list and can be loaded with the `skill` tool.
-
-The `skill` tool lets the agent load any discovered skill by name (primary, agents, or plugin).
-
-## 4. Hooks
-
-Hooks fire shell commands on lifecycle events. Configure them in JSON:
-
-- User: `~/.config/zero/hooks.json`
-- Project: `./.zero/hooks.json`
-
-```json
-{
-  "enabled": true,
-  "hooks": [
-    {
-      "id": "block-rm-rf",
-      "event": "beforeTool",
-      "matcher": "bash",
-      "command": "/usr/local/bin/zero-hook-block-rmrf.sh",
-      "enabled": true
-    },
-    {
-      "id": "log-session",
-      "event": "sessionStart",
-      "command": "/usr/local/bin/zero-hook-log.sh",
-      "enabled": true
-    }
-  ]
-}
-```
-
-The `args` array (when present) is passed verbatim to `exec.CommandContext`. The actual hook payload — event name, matcher, tool call id, tool name, tool input, tool output, status — is delivered to the command as **JSON on stdin**, not via `${...}` substitution. A typical handler reads stdin and decides what to do:
-
-```bash
-#!/usr/bin/env bash
-# /usr/local/bin/zero-hook-block-rmrf.sh
-set -euo pipefail
-payload="$(cat)"
-if printf '%s' "$payload" | grep -q '"input":"[^"]*rm[[:space:]]+-rf'; then
-  echo "refusing rm -rf" >&2
-  exit 1
-fi
-```
-
-Events the agent emits (in dispatch order):
-
-| Event | Fires when | Matcher allowed? |
-| --- | --- | --- |
-| `beforeTool` | A tool is about to run | yes (tool name) |
-| `afterTool` | A tool just returned | yes (tool name) |
-| `sessionStart` | A session begins | no |
-| `sessionEnd` | A session ends | no |
-| `specialistStart` | A sub-agent is spawned | yes (specialist name) |
-| `specialistStop` | A sub-agent ends | yes (specialist name) |
-
-A hook's exit code decides what happens next: `0` continues, non-zero blocks the tool call (`beforeTool`) or surfaces an error (`afterTool`). Hook execution is recorded in the audit log; the audit is reachable from the agent's view of past actions, not from a dedicated `zero doctor` check.
-
-> **Roadmap.** An in-UI hooks manager is on the backlog. Today you edit the JSON directly.
-
-## 5. MCP — Model Context Protocol
-
-Zero is both an **MCP client** (it can call external MCP servers) and an **MCP server** (other agents can call its tools).
-
-### As a client — configure MCP servers in `config.json`
-
-```json
-{
-  "mcp": {
-    "servers": {
-      "docs": {
-        "type": "stdio",
-        "command": "docs-mcp",
-        "args": ["--port", "7777"]
-      },
-      "github": {
-        "type": "http",
-        "url": "https://api.example.com/mcp",
-        "headers": { "Authorization": "Bearer YOUR_TOKEN_HERE" }
-      }
-    }
-  }
-}
-```
-
-Manage via CLI:
-
-```bash
-zero mcp add docs --type stdio -- docs-mcp --port 7777
-zero mcp add github --type http --url https://api.example.com/mcp \
-    --header "Authorization=Bearer YOUR_TOKEN_HERE"
-zero mcp list
-zero mcp check docs
-zero mcp remove github
-zero mcp oauth login github
-```
-
-Servers are merged from user and project configs (project wins on conflicts). Token-bearing values in `config.json` are sent verbatim — there is no `${env:...}` expansion — so prefer one of:
-
-- A wrapper script that sources the secret and execs the real command.
-- A `--header` value produced by command substitution (`"Authorization=Bearer $(print-token)"`) in a private shell config that you keep out of git.
-- A secret manager that injects the env var your MCP server reads on its own (the `command` and `args` then run inside that environment).
-
-Project MCP config remains supported for shared team servers, but a project layer
-that changes an existing server's target (`url`, `command`, or `args`) does not
-inherit user credentials. Clear or replace `headers`, `env`, or `oauth` explicitly,
-or use a new server name. OAuth tokens are bound to the resolved server identity,
-so changing a server target can require `zero mcp oauth login <server>` again.
-OAuth server names ending in `.<32 hex chars>` are reserved for token storage.
-
-### As a server — expose Zero's tools to another agent
-
-```bash
-zero serve --mcp
-zero serve --mcp -C /path/to/workspace
-zero serve --mcp --add-dir /path/to/extra
-```
-
-The server speaks MCP over stdio. Configure it from the receiving side as a `stdio` server whose command is `zero serve --mcp`.
-
-`-C/--cwd` sets the workspace root exposed as MCP resources and used by core tools. `--add-dir` (repeatable) widens that resource/tool scope beyond the workspace without granting the sandbox temp roots used by interactive runs.
-## 6. Plugins
-
-A plugin is a self-contained directory that bundles tools, hooks, and skills for one capability. Plugins live at:
-
-- User: `~/.config/zero/plugins/<id>/`
-- Project: `./.zero/plugins/<id>/`
-
-Each plugin has a `plugin.json` manifest:
-
-```json
-{
-  "id": "github-pr-review",
-  "name": "GitHub PR Review",
-  "description": "Adds review tools for GitHub PRs.",
-  "version": "1.0.0",
-  "tools": [
-    { "name": "list_prs", "command": "./tools/list_prs.sh" }
-  ],
-  "hooks": [
-    { "name": "pre-merge-check", "event": "beforeTool", "command": "./hooks/pre-merge.sh" }
-  ],
-  "skills": [
-    { "path": "./skills/review-checklist/SKILL.md" }
-  ]
-}
-```
-
-Install and manage:
-
-```bash
-zero plugins add ./github-pr-review      # copy into ~/.config/zero/plugins/ or ./.zero/plugins/
-zero plugins list
-zero plugins info github-pr-review       # show manifest path, enabled state, lock metadata
-zero plugins remove github-pr-review    # alias: rm
-```
-
-A plugin is enabled by being present in the plugins directory and disabled by removing it (or by the user setting `"enabled": false` in its `plugin.json`). Plugins are not enabled or disabled by a CLI subcommand today.
-
-Plugin commands run with the plugin directory as their working directory. Use relative paths; the loader resolves them at activation time.
-
-> **Roadmap.** An in-UI plugins manager (browse, install, enable / disable) is on the backlog. Today you use the `zero plugins` CLI subcommands above.
-
-## 7. Configuration locations
-
-Five configuration sources, in precedence order (later sources override earlier ones):
-
-| Source | Path / Key | Notes |
-| --- | --- | --- |
-| Built-in defaults | compiled in | Lowest priority. |
-| User config | `~/.config/zero/config.json` | Your machine. Never committed. |
-| Project config | `./.zero/config.json` | The repo. Committed (or not, your call). |
-| Environment | `ZERO_*` | Provider commands, secrets, skills dir override. |
-| CLI flags | `--model`, `--mode`, ... | Highest priority, per-invocation. |
-
-The user config holds things that should follow the user across projects (default provider, default model, theme). The project config holds things the team agreed on (provider catalog, sandbox policies, model restrictions).
-
-The sandbox `additionalWriteRoots` key is **ignored in project config** by design — a checked-out repo cannot widen its own sandbox. Set it in the user config or pass `--add-dir` per-invocation.
-
-## 8. End-to-end example
-
-A team that wants every contributor's Zero to behave the same way commits:
-
-- `AGENTS.md` — project conventions, build commands, do-not-edit lists.
-- `.zero/config.json` — provider catalog, default model, allowed tools.
-- `.zero/specialists/api-reviewer.md` — the team's PR-review specialist.
-- `.zero/hooks.json` — block `rm -rf` on `beforeTool`.
-- `.zero/plugins/internal-tooling/` — a plugin that adds the team's internal CLI tools to the agent's toolset.
-
-Each contributor adds only:
-
-- `~/.config/zero/config.json` — their personal API keys, theme, default mode.
-- `~/.config/zero/ZERO.md` — personal preferences that follow them across every project (see section 1).
-- `~/.local/share/zero/skills/` — personal skills they keep across projects.
-
-That's it. Run `zero` from the repo root and the agent has the team's full instruction set, every contributor's personal setup, and nothing else.
-
-## 9. Reference
-
-- [README](../README.md) — install, quickstart, command reference.
-- [docs/SPECIALISTS.md](SPECIALISTS.md) — full specialist manifest spec.
-- [docs/STREAM_JSON_PROTOCOL.md](STREAM_JSON_PROTOCOL.md) — `zero exec` I/O contract.
-- [docs/INSTALL.md](INSTALL.md) — install from source or release.

docs/HOW_ZERO_WORKS.md

@@ -1,801 +0,0 @@
-# How Zero Works
-
-This document explains Zero's architecture at a high level: the layers, module
-boundaries, data flow, and runtime concepts behind the interactive TUI and the
-headless `zero exec` mode.
-
-It is written as a progressive map: each high-level section introduces the
-concepts you need for orientation, then points to the later section that explains
-the mechanics in more detail.
-
-## Big Picture
-
-Zero is a Go CLI application with two main user-facing surfaces:
-
-- **Interactive TUI** (`zero`): a Bubble Tea application that renders chat,
-  tool cards, permission prompts, model/provider pickers, and session controls.
-- **Headless exec** (`zero exec ...`): a scriptable command runner that uses the
-  same agent loop and tools, but writes text, JSON, or stream-JSON events.
-
-Both surfaces converge on the same core agent runtime:
-
-```mermaid
-flowchart TD
-    User[User / automation] --> CLI[cmd/zero -> internal/cli]
-
-    CLI --> TUI[Interactive TUI<br/>internal/tui]
-    CLI --> Exec[Headless exec<br/>internal/cli/exec.go]
-    CLI --> ACP[ACP / editor bridge<br/>internal/acp]
-
-    TUI --> Agent[Agent loop<br/>internal/agent]
-    Exec --> Agent
-    ACP --> Agent
-
-    Agent --> Provider[Provider runtime<br/>internal/zeroruntime.Provider]
-    Provider --> APIs[OpenAI / Anthropic / Gemini / compatible APIs]
-
-    Agent --> Registry[Tool registry<br/>internal/tools.Registry]
-    Registry --> CoreTools[Core tools<br/>read/edit/shell/search/plan]
-    Registry --> MCP[MCP tools<br/>internal/mcp]
-    Registry --> Specialists[Specialist tools<br/>internal/specialist]
-    Registry --> Plugins[Plugin tools<br/>internal/plugins]
-    Registry --> LocalControl[Browser / terminal / desktop helpers]
-
-    Agent --> Sandbox[Permission + sandbox engine<br/>internal/sandbox]
-    Agent --> Hooks[Hooks dispatcher<br/>internal/hooks]
-    Agent --> Sessions[Session event store<br/>internal/sessions]
-```
-
-The important split is:
-
-- `internal/tui` is the **presentation layer**. For the mechanics, see
-  [Presentation Layer: Bubble Tea TUI](#presentation-layer-bubble-tea-tui) and
-  [How the TUI and Agent Loop Connect](#how-the-tui-and-agent-loop-connect).
-- `internal/agent` is the **reasoning/tool loop**. For the per-turn mechanics,
-  see [Agent Loop](#agent-loop), [Tool Execution Lifecycle](#tool-execution-lifecycle),
-  and [Context Budget and Compaction](#context-budget-and-compaction).
-- `internal/tools`, `internal/sandbox`, and provider packages are the **runtime
-  capability layer**. For the details, see [Runtime Provider Layer](#runtime-provider-layer),
-  [Tools and Capabilities](#tools-and-capabilities), and
-  [Permission and Sandbox Model](#permission-and-sandbox-model).
-- `internal/sessions` is the **durable event log** used for resume, fork,
-  rewind, sub-agents, and history. See [Session Storage and History](#session-storage-and-history).
-
-## Architecture Spine
-
-The easiest way to understand Zero is not as a set of independent packages, but
-as one spine with several mechanisms attached to it:
-
-```mermaid
-flowchart LR
-    Surface[Surface<br/>TUI / exec / ACP] --> Run[Run request<br/>prompt + runtime options]
-    Run --> Agent[Agent loop<br/>turns, tools, recovery]
-    Agent --> Capabilities[Capabilities<br/>providers + tools + sandbox]
-    Capabilities --> Events[Runtime events<br/>text, tool calls, permissions, usage]
-    Events --> Surface
-    Events --> Sessions[Session log<br/>resume, fork, rewind, history]
-
-    Config[Config / guidelines / skills] --> Run
-    Extensions[MCP / specialists / plugins / hooks] --> Capabilities
-```
-
-The **TUI + agent loop** is the main story because it explains the normal user
-experience: type a prompt, watch the model stream, approve tool use, see file
-changes, and receive a final answer. Other mechanisms are still important, but
-most of them attach to the same spine:
-
-| Mechanism | Where it attaches | What it changes |
-| --- | --- | --- |
-| Config, provider profiles, model registry | Before a run starts | Which provider/model is used and what defaults apply. |
-| Project/user guidelines, skills, specialists context | Prompt construction | What instructions and available capabilities the model sees. |
-| Tool registry, MCP, plugins, local-control helpers | Capability layer | Which actions the model may request. |
-| Permission policy, sandbox, command-prefix grants | Tool execution | Whether a requested action can run, needs approval, or is denied. |
-| Hooks | Tool lifecycle | Optional before/after observers or blockers around tool calls. |
-| Sessions, checkpoints, compaction metadata | Event persistence/context | What can be resumed, forked, rewound, or summarized for later turns. |
-| Stream-JSON | Surface layer | How the same runtime events are emitted for automation instead of rendered in the TUI. |
-
-So this document focuses first on the TUI/agent-loop bridge, then explains the
-supporting mechanisms in terms of where they connect to that bridge.
-
-Use this as the reading path:
-
-| If you see this concept in a high-level flow | Read this detail section |
-| --- | --- |
-| Surface submits a run | [Startup Flow](#startup-flow), then [How the TUI and Agent Loop Connect](#how-the-tui-and-agent-loop-connect) |
-| Agent builds context and advances turns | [Agent Loop](#agent-loop) |
-| “Maybe compact context” | [Context Budget and Compaction](#context-budget-and-compaction) |
-| Model requests a tool | [Tool Execution Lifecycle](#tool-execution-lifecycle), then [Tools and Capabilities](#tools-and-capabilities) |
-| Tool needs approval or sandbox access | [Permission and Sandbox Model](#permission-and-sandbox-model) |
-| Events appear in the UI or stream-JSON | [How the TUI and Agent Loop Connect](#how-the-tui-and-agent-loop-connect), [Stream-JSON Protocol](#stream-json-protocol) |
-| Resume, fork, rewind, or sub-agent history | [Session Storage and History](#session-storage-and-history) |
-
-## Main Packages
-
-The sections above describe the architecture in concepts. If you want to connect
-those concepts to implementation details, use this table as the code map before
-reading the deeper flows below.
-
-| Package | Role |
-| --- | --- |
-| `cmd/zero` | Tiny binary entrypoint; calls `internal/cli.Run`. |
-| `internal/cli` | Command parsing, config resolution, provider creation, registry setup, TUI/exec/ACP launch paths. |
-| `internal/tui` | Bubble Tea model/update/view code, transcript rendering, composer, prompts, slash commands, setup wizard. |
-| `internal/agent` | Core loop that sends messages to the model, streams output, executes tools, handles permissions, compaction, retries, and final answers. |
-| `internal/zeroruntime` | Provider-neutral message, tool-call, stream, usage, and image types. Providers implement `Provider.StreamCompletion`. |
-| `internal/providers` | Factory that turns a resolved provider profile into an OpenAI, Anthropic, Gemini, Codex, or compatible provider adapter. |
-| `internal/tools` | Tool interface, registry, core read/write/shell/network tools, local-control wrappers, tool safety metadata, redaction, output ceilings. |
-| `internal/sandbox` | Permission policy, path/network checks, grant store, command-prefix approvals, and platform sandbox backends. |
-| `internal/sessions` | Local append-only session storage: `metadata.json` + `events.jsonl`. |
-| `internal/config` | User/project config resolution, active provider, preferences, sandbox/tool settings. |
-| `internal/mcp` | MCP server config, client runtime, permission store, and MCP tool registration. |
-| `internal/specialist` / `internal/swarm` | Sub-agent manifests and team/member orchestration exposed as tools. |
-| `internal/localcontrol` / `internal/browser` | Runtime helpers used by config-gated local-control tool wrappers. |
-| `internal/plugins` / `internal/skills` / `internal/hooks` | Extension surfaces loaded into the agent context, tool registry, or tool lifecycle. |
-| `internal/streamjson` | Machine-readable protocol for `zero exec --input-format/--output-format stream-json`. |
-
-## Startup Flow
-
-### Interactive TUI (`zero`)
-
-```mermaid
-sequenceDiagram
-    participant Main as cmd/zero
-    participant CLI as internal/cli
-    participant Config as internal/config
-    participant TUI as internal/tui
-    participant Agent as internal/agent
-
-    Main->>CLI: Run(args)
-    CLI->>Config: resolve config for workspace
-    CLI->>CLI: build provider
-    CLI->>CLI: build tool registry
-    CLI->>CLI: register core, config-gated local-control, specialists, MCP, plugins
-    CLI->>CLI: create sandbox engine + session store
-    CLI->>TUI: run Bubble Tea with Options
-    TUI->>TUI: Init / Update / View event loop
-    TUI->>Agent: agent.Run on submitted prompt
-```
-
-The TUI owns terminal interaction. It tracks composer state, transcript rows,
-scrolling, permission dialogs, model/provider pickers, setup wizard state, and
-visual updates. When the user submits a prompt, the TUI builds `agent.Options`
-with callbacks like `OnText`, `OnToolCall`, `OnToolResult`, `OnPermission`,
-`OnAskUser`, and `OnUsage`. Those callbacks convert runtime events into visible
-transcript rows and session events.
-
-### Headless exec (`zero exec`)
-
-```mermaid
-flowchart TD
-    A[Parse exec flags] --> B[Resolve config + provider]
-    B --> C[Build tool registry]
-    C --> D[Resolve prompt / stream-json input / images]
-    D --> E[Prepare or resume session]
-    E --> F[Create stream-json/text/json writer]
-    F --> G[agent.Run]
-    G --> H[Write final output and exit code]
-```
-
-`zero exec` uses the same `agent.Run` loop as the TUI, but without interactive
-terminal state. It can persist sessions, resume/fork existing sessions, run in
-stream-JSON mode for automation, and always requires a stronger completion
-signal so a no-tool assistant response is not treated as success while work is
-still pending.
-
-## Agent Loop
-
-The core loop lives in `internal/agent.Run`. Conceptually, every turn follows
-this cycle:
-
-```mermaid
-flowchart TD
-    Start[Build system prompt + user prompt] --> Expose[Partition visible tools]
-    Expose --> Compact[Maybe compact context]
-    Compact --> Request[Send CompletionRequest to provider]
-    Request --> Stream[Collect streamed text, reasoning, usage, tool calls]
-
-    Stream --> HasTools{Tool calls?}
-    HasTools -- No --> FinalCheck{Final answer acceptable?}
-    FinalCheck -- Yes --> Done[Return Result]
-    FinalCheck -- No --> Nudge[Append user nudge and continue]
-
-    HasTools -- Yes --> Execute[Execute tool calls]
-    Execute --> Permission[Permission / sandbox / hooks]
-    Permission --> ToolResult[Append tool result messages]
-    ToolResult --> Diagnostics[Post-edit diagnostics / self-correct]
-    Diagnostics --> Expose
-```
-
-Key responsibilities of the loop:
-
-1. **Build prompt state** — combines the system prompt, project/user guidelines,
-   loaded skills/specialist instructions, the current user prompt, and images. The
-   resulting message list is the model-facing context that later tool results are
-   appended to. Resumed session history is replayed by the caller before a new run,
-   not passed as a separate `agent.Run` history field.
-2. **Expose tools** — starts from the registry built by the CLI, then filters
-   tool definitions by permission mode, enabled or disabled filters, and
-   deferred-loading rules. Tool schemas also count toward context budget, so they
-   matter for compaction decisions.
-3. **Call provider** — sends a provider-neutral `zeroruntime.CompletionRequest`.
-4. **Stream events** — forwards text, reasoning, tool-call deltas, and usage to
-   the caller via callbacks.
-5. **Execute tools** — decodes tool arguments, checks filters, applies
-   permissions and sandbox policy, dispatches hooks, runs the registered tool,
-   redacts output, and appends a tool message. See
-   [Tool Execution Lifecycle](#tool-execution-lifecycle).
-6. **Recover and guard** — handles malformed tool calls, repeated failures,
-   empty turns, context pressure, stream reconnect/retry, and max-turn fallback.
-   Context pressure is handled by [Context Budget and Compaction](#context-budget-and-compaction).
-7. **Finalize** — returns a final answer or a structured incomplete/stop reason.
-
-The model never directly mutates the workspace. It asks for tool calls; Zero
-validates and executes those calls through the registry, permission policy, and
-sandbox engine.
-
-## Tool Execution Lifecycle
-
-A tool call is part of the model conversation, not a side channel that bypasses
-it. The model proposes an action, Zero decides whether that action is allowed to
-run, the tool returns a structured result, and that result becomes context for the
-next provider request.
-
-```mermaid
-sequenceDiagram
-    participant Model as LLM / provider
-    participant Agent as agent.Run
-    participant Policy as filters + permissions + sandbox
-    participant Hooks as hooks
-    participant Tool as tool implementation
-    participant Surface as TUI / exec callbacks
-    participant Transcript as model conversation + session events
-
-    Agent->>Model: request with messages + available tool schemas
-    Model-->>Agent: assistant text and/or tool_call(name, args)
-    Agent-->>Surface: stream text / tool-call preview
-    Agent->>Agent: collect and decode tool arguments
-    Agent->>Policy: check tool filters, permission mode, sandbox policy
-
-    alt denied before execution
-        Policy-->>Agent: denied result
-        Agent-->>Surface: tool failed/denied event
-        Agent->>Transcript: append tool result explaining denial
-    else needs user approval
-        Policy-->>Surface: permission request
-        Surface-->>Policy: allow / deny / remember decision
-        Policy-->>Agent: decision
-    end
-
-    alt approved to execute
-        Agent->>Hooks: beforeTool hooks
-        Hooks-->>Agent: allow / block / annotate
-        Agent->>Tool: execute(ctx, args)
-        Tool-->>Agent: output, error, metadata
-        Agent->>Hooks: afterTool hooks
-        Agent->>Agent: redact secrets, apply semantic output budget, enforce byte ceiling
-        Agent->>Agent: classify success/failure
-        Agent-->>Surface: OnToolResult callback
-        Agent->>Transcript: append tool result message
-    end
-
-    Agent->>Model: next request includes prior assistant tool_call + tool result
-    Model-->>Agent: continue with more tools or final answer
-```
-
-Conceptually, one requested tool goes through these phases:
-
-1. **Proposal** — the provider streams assistant content and may emit one or more
-   tool calls. Zero can show the partial tool arguments in the TUI while they are
-   still streaming, but nothing has executed yet.
-2. **Decode and gate** — the agent validates JSON arguments, checks whether the
-   tool is enabled, applies permission mode, asks the sandbox engine about path,
-   network, shell, or command-prefix constraints, and may invoke permission
-   callbacks.
-3. **Execute** — the selected registry entry runs locally, through MCP, as a
-   specialist/sub-agent, or through a plugin/local-control helper. Hooks can run
-   before or after this step.
-4. **Normalize result** — Zero turns the outcome into an agent `ToolResult`: a
-   success, failure, denial, cancellation, timeout, or validation error with
-   output/metadata that may be redacted or truncated.
-5. **Append to context** — the result is appended as a tool-result message paired
-   with the assistant tool call. The next model request includes that result, so
-   the model can inspect it and decide what to do next.
-6. **Render and persist** — the result is surfaced to the TUI, stream-JSON
-   writer, or ACP client. TUI and exec record rich tool/session events for
-   resume/history; ACP currently persists conversational messages and streams
-   tool/permission updates to the client.
-
-Tool success and failure are both informative context. A successful `read_file`
-result gives the model file contents; a failed `edit_file` result gives the model
-an error such as “old string not found”; a denied shell command tells the model
-that the action is not currently permitted. The model is expected to use that
-feedback to continue, choose another tool, ask the user, or explain the blocker.
-
-The judgement that “information is sufficient” is made by the model, not by the
-TUI. Zero controls the loop mechanics: it keeps appending messages, enforces
-safety, retries or nudges around malformed/empty/incomplete turns, and stops on
-hard limits. But the semantic decision to call another tool versus produce a
-final answer comes from the next provider response. In other words:
-
-```mermaid
-flowchart TD
-    Result[Tool result appended to conversation] --> Next[Next provider request]
-    Next --> Decide{Model response}
-    Decide -- More information/action needed --> MoreTools[Request another tool]
-    MoreTools --> Result
-    Decide -- User input needed --> AskUser[Call ask_user or explain question]
-    AskUser --> Result
-    Decide -- Enough information --> Final[Final assistant answer]
-    Decide -- Invalid / empty / unsafe --> Guard[Agent guardrail, retry, denial, or stop reason]
-    Guard --> Next
-```
-
-This is why tool output is usually more than UI decoration. It is part of the
-working memory for the next model turn, subject to context-window pressure and
-compaction. The TUI may summarize or prettify it for the user, but the agent loop
-also stores the machine-facing result that the model sees.
-
-## Context Budget and Compaction
-
-The `Maybe compact context` step in the agent-loop diagram is Zero's context-window
-pressure valve. It is not a separate memory engine; it rewrites the current
-model-facing conversation so the next provider request fits within the configured
-context window.
-
-Compaction is enabled only when the run knows a positive `ContextWindow`. When it
-is enabled, the loop does this near the top of each turn:
-
-```mermaid
-flowchart TD
-    Tools[Partition visible tools] --> Estimate[Estimate messages + tool-definition tokens]
-    Estimate --> Under{Under threshold?}
-    Under -- Yes --> Keep[Keep context as-is]
-    Under -- No --> Grown{Grown past low-water mark?}
-    Grown -- No --> Keep
-    Grown -- Yes --> Prune[Prune stale large tool outputs]
-    Prune --> Recheck{Back under threshold?}
-    Recheck -- Yes --> KeepPruned[Use pruned context]
-    Recheck -- No --> Summarize[Summarize oldest middle]
-    Summarize --> Replace[system messages + summary + recent suffix]
-```
-
-The trigger is intentionally approximate:
-
-- Zero estimates request size using message text, images, tool-call arguments, and
-  the tool definitions advertised on that turn.
-- The proactive threshold is roughly **two-thirds of the context window**.
-- Provider usage from completed turns is used to calibrate the rough estimate over
-  time, so the budget gauge and compaction decision track the provider more
-  closely after a few turns.
-- A low-water mark prevents the loop from re-summarizing the same already-compacted
-  history on every turn.
-
-When compaction actually runs, it tries to preserve the parts most likely to be
-needed by the model:
-
-- leading system messages stay verbatim;
-- the most recent turns stay verbatim (`CompactionPreserveLast`, defaulting to the
-  agent's normal preserve count);
-- the preserved suffix is widened to a provider-valid turn boundary so tool-result
-  messages are not replayed without their corresponding assistant tool call;
-- the older middle is summarized into one user-role message labeled
-  `[Summary of earlier conversation]`;
-- structured state that should not be paraphrased away, such as active plan/loaded
-  skill state, is preserved alongside the prose summary.
-
-There are two compaction paths:
-
-1. **Proactive compaction** runs before the provider call when the estimated
-   request is over threshold.
-2. **Reactive compaction** runs after a provider returns a context-limit style
-   error. Zero compacts once and retries that same turn; if the provider still
-   rejects the request, the error is surfaced rather than looping forever.
-
-Compaction failures are conservative. If the summarizer fails during proactive
-compaction, Zero keeps the original messages and tries to continue. It does not
-drop conversation history just to make room.
-
-## Presentation Layer: Bubble Tea TUI
-
-`internal/tui` is a Bubble Tea application. The usual Bubble Tea shape applies:
-
-```mermaid
-flowchart LR
-    Init[model.Init] --> Update[model.Update]
-    Update --> View[model.View]
-    View --> Terminal[Terminal frame]
-    Terminal --> Msg[tea.Msg: key, mouse, window, agent event]
-    Msg --> Update
-```
-
-The TUI is intentionally not the agent brain. It is responsible for:
-
-- collecting user input from the composer;
-- rendering assistant text, reasoning, tool cards, permissions, plans, usage,
-  files, and sidebars;
-- handling slash commands and modal surfaces;
-- converting agent callbacks into Bubble Tea messages;
-- persisting visible session events;
-- allowing live provider/model/permission changes.
-
-When a prompt is submitted, the TUI starts a background command that invokes
-`agent.Run`. Streaming callbacks send messages back into the Bubble Tea update
-loop, so rendering remains responsive while the agent is working.
-
-## How the TUI and Agent Loop Connect
-
-A common source of confusion is that Zero has two event loops running at once:
-
-1. **Bubble Tea loop** — owns terminal input, in-memory UI state, and frames.
-2. **Agent loop** — owns model turns, tool execution, permissions, retries, and
-   stop conditions.
-
-They are connected by `tea.Cmd` and `agent.Options` callbacks. The TUI never
-runs provider/tool logic directly inside `Update`; instead, submit handling gates
-and dispatches a prompt, the TUI starts a background command, and that command
-calls `agent.Run` with callbacks that send runtime events back to the Bubble Tea
-program. Function names in this section are illustrative code anchors, not API
-boundaries.
-
-```mermaid
-sequenceDiagram
-    participant User
-    participant TUI as Bubble Tea model<br/>internal/tui
-    participant Cmd as tea.Cmd goroutine<br/>runAgentWithOptions
-    participant Agent as agent.Run<br/>internal/agent
-    participant Provider as Provider.StreamCompletion
-    participant Tools as Tool registry + sandbox
-    participant Store as Session store
-
-    User->>TUI: type prompt / press Enter
-    TUI->>TUI: validate/gate submit and append visible user row
-    TUI->>Cmd: launch tea.Cmd with runID + cancellable context
-    Cmd->>Agent: agent.Run(prompt, provider, options)
-    Agent->>Provider: stream completion request
-    Provider-->>Agent: text / reasoning / tool-call deltas
-    Agent-->>Cmd: OnText / OnReasoning / OnToolCallStart callbacks
-    Cmd-->>TUI: runtimeMessageSink(tea.Msg)
-    TUI->>TUI: Update appends/updates transcript rows
-    TUI-->>User: View renders new frame
-
-    Agent->>Tools: execute collected tool call
-    Tools->>Tools: permission, sandbox, hooks, redaction
-    Tools-->>Agent: tool result
-    Agent-->>Cmd: OnToolCall / OnToolResult / OnPermission / OnUsage
-    Cmd->>Store: batch session events for resume/history
-    Cmd-->>TUI: final agentResponseMsg
-    TUI->>TUI: mark run complete, flush rows/events, unlock composer
-```
-
-In the TUI path, the same event can have three representations:
-
-| Runtime thing | Agent representation | TUI representation | Session representation |
-| --- | --- | --- | --- |
-| Assistant text | streamed provider text forwarded through `OnText` | incremental assistant row text | assistant message event |
-| Reasoning preview | `OnReasoning` callback | collapsible/preview reasoning row | usually not the same as final assistant text |
-| Tool call | `agent.ToolCall` from collected provider stream | tool card, specialist card, or plan panel update | `EventToolCall` |
-| Tool result | `agent.ToolResult` | result card/status, changed files, plan/specialist updates | `EventToolResult` |
-| Permission prompt/audit | `PermissionRequest` / `PermissionEvent` | modal prompt or permission row | permission request/decision events |
-| Usage/context | provider usage + measured context | context gauge / usage metadata | usage event |
-
-### One interactive turn, end to end
-
-For a normal prompt in the interactive TUI:
-
-1. Submit handling validates composer state and command gates, then dispatches
-   the prompt to the launch path.
-2. The launch path appends immediate UI rows, creates a cancellable run context,
-   and starts `runAgent` as a `tea.Cmd`; run bookkeeping assigns the `runID` used
-   to correlate later events.
-3. `runAgentWithOptions` copies the current UI-controlled runtime choices into
-   `agent.Options`: provider/model labels, permission mode, cwd, session ID,
-   images, registry, response style, context window, and self-correction hooks.
-4. The command wraps agent callbacks. For example:
-   - `OnText` sends incremental assistant text into the TUI message sink.
-   - `OnToolCallStart` / `OnToolCallDelta` show live tool-argument streaming.
-   - `OnToolCall` creates visible tool rows and appends session events.
-   - `OnToolResult` creates result rows, updates plan/specialist side panels, and
-     records changed files.
-   - `OnPermissionRequest` sends a prompt to the TUI and blocks the agent loop
-     until the user answers or the run context is cancelled.
-   - `OnAskUser` does the same for model-initiated clarification questions.
-5. `agent.Run` builds provider-neutral messages and tool definitions, streams a
-   provider response, executes any requested tools through the registry/sandbox,
-   appends tool results back into the model conversation, and repeats until it has
-   a final answer or a stop condition.
-6. When `agent.Run` returns, the command emits `agentResponseMsg`. The Bubble Tea
-   loop marks the run complete, flushes queued rows/session events, updates usage
-   and file state, and re-enables the composer.
-
-The important architectural point: **the TUI is stateful and reactive, but the
-agent loop is the source of runtime truth for model/tool progress**. Presentation
-state is derived from callback events and persisted session events; workspace
-changes happen only through tools executed by the agent runtime.
-
-### Feature flows use the same bridge
-
-For one interactive turn, the supporting systems do not run beside the TUI/agent
-loop; they run inside or around it:
-
-```mermaid
-flowchart TD
-    Submit[TUI submit] --> Options[Build agent.Options]
-    Options --> Context[Assemble context<br/>config, guidelines, history, skills]
-    Context --> ProviderCall[Provider stream]
-    ProviderCall --> StreamEvents[Text/reasoning/tool-call events]
-    StreamEvents --> Render[TUI render callbacks]
-
-    ProviderCall --> ToolCall{Tool call?}
-    ToolCall -- No --> Final[Final response]
-    ToolCall -- Yes --> ToolGate[Tool gate<br/>filters, permissions, sandbox]
-    ToolGate --> HooksBefore[beforeTool hooks]
-    HooksBefore --> Execute[Tool / MCP / specialist / plugin]
-    Execute --> HooksAfter[afterTool hooks]
-    HooksAfter --> ToolResult[Tool result message]
-    ToolResult --> Persist[Session events]
-    ToolResult --> Context
-
-    Final --> Persist
-    Persist --> Resume[Resume / fork / rewind / parent metadata]
-```
-
-Most interactive features are variations of that same bridge rather than separate
-architectures:
-
-- **Slash commands** either mutate TUI state directly (`/model`, `/permission`,
-  `/help`) or start short `tea.Cmd` jobs that later return a message to `Update`
-  (`/doctor`, `/search`, `/compact`, session commands).
-- **Permissions** sit between a model-requested tool call and actual execution:
-  the agent blocks on `OnPermissionRequest`, the TUI surfaces a modal and sends
-  the decision back through a channel, and the session log records the
-  request/decision events.
-- **Plans, specialists/sub-agents, and ask-user prompts** all follow the same
-  runtime pattern: the model requests a capability, the agent records the
-  call/result in its conversation, and the TUI mirrors the visible state as plan
-  panels, specialist cards, or question forms. Specialist session metadata can
-  carry parent/root and sub-agent/spec fields, with richer linkage depending on
-  the specialist or swarm path.
-- **Streaming tool writes** start as provider tool-call deltas before execution;
-  the TUI shows the arguments live, then replaces/augments them with the actual
-  `ToolResult` after the tool runs.
-- **Compaction** is an agent-loop context-management step; the TUI may show that
-  it happened, but the purpose is to keep the next provider request within budget.
-- **Stream-JSON** is another surface over the same callbacks: `zero exec`
-  serializes the events as JSON protocol events instead of Bubble Tea views.
-
-So a feature-specific deep dive should usually explain: what user event starts it,
-which `tea.Msg`/callback carries it, which agent/tool event is authoritative, and
-which session events make it resumable.
-
-## Runtime Provider Layer
-
-Providers are hidden behind this interface in `internal/zeroruntime`:
-
-```go
-type Provider interface {
-    StreamCompletion(ctx context.Context, request CompletionRequest) (<-chan StreamEvent, error)
-}
-```
-
-The agent loop only knows about provider-neutral messages, tool definitions,
-stream events, reasoning blocks, images, and token usage. Provider adapters in
-`internal/providers/*` translate those neutral types to provider-specific APIs:
-
-- OpenAI and OpenAI-compatible APIs;
-- Anthropic and Anthropic-compatible APIs;
-- Gemini;
-- Codex / ChatGPT OAuth-backed runtime.
-
-`internal/providers.New` resolves the configured model through the model
-registry/catalog, normalizes API model IDs and base URLs, attaches API keys or
-OAuth resolvers, and returns the concrete adapter.
-
-## Tools and Capabilities
-
-Tools implement a common interface and are registered by name in
-`internal/tools.Registry`.
-
-```mermaid
-flowchart TD
-    AgentToolCall[Tool call from model] --> Decode[Decode JSON args]
-    Decode --> Filter[Enabled/disabled + permission-mode filter]
-    Filter --> Precheck[Pre-permission rejecters]
-    Precheck --> SandboxEval[Sandbox Evaluate]
-    SandboxEval --> Prompt{Needs user approval?}
-    Prompt -- Yes --> Decision[Permission callback]
-    Prompt -- No --> Run[Registry.RunWithOptions]
-    Decision --> Run
-    Run --> Redact[Redact secrets]
-    Redact --> Budget[Token-aware semantic output budget]
-    Budget --> Ceiling[Existing hard byte ceiling]
-    Ceiling --> Message[Return one tool result to model]
-```
-
-Oversized results use deterministic, provider-neutral estimated-token budgets.
-Policies retain useful structure for files, search matches, tests, process logs,
-diffs, and worker conclusions; tools without a declared category use a UTF-8-safe
-head/tail fallback. The estimate is intentionally conservative for non-ASCII
-text and is not exact provider tokenization. Existing byte ceilings remain the
-authoritative safety limit. When the existing spill mechanism can persist the
-complete redacted text received by the budgeting layer, the result includes its
-safe spill reference. This does not imply capture of subprocess bytes already
-discarded by a tool's established internal buffer.
-
-Core tool groups include:
-
-- **Read-only tools**: file reads, directory listing, glob, grep, LSP navigation,
-  skills, `ask_user`, permission requests.
-- **Write tools**: `write_file`, `edit_file`, `apply_patch`, `update_plan`.
-- **Shell tools**: `exec_command`, `write_stdin`, legacy `bash`.
-- **Network tools**: `web_fetch` and optional web search backend.
-- **Local-control tools**: config-gated wrappers for browser, terminal, desktop,
-  and artifact helpers backed by packages such as `internal/localcontrol` and
-  `internal/browser`.
-- **Extension tools**: MCP tools, specialists/sub-agents, swarm tools, plugins.
-
-Tool definitions include safety metadata: side-effect class, permission level,
-reason, and whether prompt-gated tools are advertised in auto-like modes. The
-agent uses this metadata plus per-turn filters and deferred-loading state to
-choose which registered tools the model can see and whether a call can run
-immediately.
-
-## Permission and Sandbox Model
-
-Zero has two related but separate checks:
-
-1. **Tool permission metadata** says whether a tool is allowed, denied, or needs
-   approval.
-2. **Sandbox policy** evaluates the actual request: path scope, network access,
-   destructive shell patterns, explicit escalation, persistent/session grants,
-   and platform isolation availability.
-
-```mermaid
-flowchart TD
-    Call[Tool call] --> Safety[Tool Safety metadata]
-    Call --> Scope[Workspace/additional roots scope]
-    Safety --> Engine[Sandbox Engine Evaluate]
-    Scope --> Engine
-    Engine --> Allow[Allow]
-    Engine --> Prompt[Prompt user]
-    Engine --> Deny[Deny]
-    Prompt --> Grant[Turn/session/persistent grant]
-    Grant --> Execute[Execute tool]
-    Allow --> Execute
-    Deny --> Result[Return denied tool result]
-```
-
-The sandbox engine can:
-
-- auto-allow workspace writes for safe file-edit tools;
-- block reads/writes outside the allowed scope;
-- prompt for recoverable outside-workspace access;
-- deny or prompt for network depending on policy;
-- detect destructive shell commands;
-- run shell commands through a platform backend when available;
-- remember session or persistent grants;
-- support narrow command-prefix approvals.
-
-The user-facing TUI and `zero exec` mode both route permission requests through
-callbacks, but the TUI can show interactive modals while headless mode generally
-returns non-interactive fallback behavior unless a protocol/client handles it.
-
-## Session Storage and History
-
-Sessions live locally under Zero's data directory. A session directory contains:
-
-- `metadata.json` — session ID, title, cwd, provider/model, parent/root IDs,
-  spec/sub-agent fields, timestamps, and event count.
-- `events.jsonl` — append-only event log with messages, tool calls/results,
-  permissions, usage, checkpoints, forks, compaction, and specs. TUI and exec use
-  the richer event stream; ACP currently records conversational message history
-  while streaming tool and permission updates to the client.
-
-```mermaid
-flowchart LR
-    TUI[TUI callbacks] --> Store[internal/sessions.Store]
-    Exec[exec recorder] --> Store
-    Store --> Metadata[metadata.json]
-    Store --> Events[events.jsonl]
-    Events --> Resume[resume]
-    Events --> Fork[fork]
-    Events --> Rewind[rewind/checkpoints]
-    Events --> Children[parent/root metadata]
-```
-
-This is why sessions can be resumed, forked, rewound, and inspected without a
-remote service. Zero stores the local event history; providers only receive the
-messages needed for the current completion request.
-
-## Stream-JSON Protocol
-
-`zero exec --output-format stream-json` emits newline-delimited JSON events for
-integrations and CI. The protocol is implemented in `internal/streamjson` and is
-specified in [STREAM_JSON_PROTOCOL.md](STREAM_JSON_PROTOCOL.md).
-
-Common output event types include:
-
-- `run_start`, `text`, `reasoning`, `tool_call`, `tool_result`;
-- `permission`, `permission_request`, `permission_decision`;
-- `checkpoint`, `restore`, `usage`, `final`, `warning`, `error`, `run_end`.
-
-Input stream-JSON can provide `prompt` or `message` events and image payloads.
-The CLI parses those events, resolves a single prompt, validates/decodes images,
-and passes the result into the same agent loop used everywhere else.
-
-## Extensibility Flow
-
-Zero's extension surfaces are loaded before the agent loop starts:
-
-```mermaid
-flowchart TD
-    Config[Config resolution] --> Registry[Tool registry]
-    Registry --> Core[Core tools]
-    Registry --> MCP[MCP server tools]
-    Registry --> Specialist[Specialist / Task tools]
-    Registry --> Plugin[Plugin tools]
-
-    Config --> Prompt[System prompt context]
-    Prompt --> Guidelines[AGENTS.md / ZERO.md]
-    Prompt --> Skills[Skills list]
-    Prompt --> SpecialistContext[Specialist descriptions]
-
-    ToolLifecycle[Tool lifecycle] --> Hooks[beforeTool / afterTool hooks]
-```
-
-- **Project/user guidelines** become prompt context.
-- **Skills** are instruction packs the model can load on demand.
-- **Specialists** are sub-agents callable through the `Task` tool.
-- **MCP servers** contribute external tools.
-- **Plugins** can add tools, hooks, and skill roots. Bootstrap always registers a
-  multi-root skill tool: primary Zero skills dir, optional `~/.agents/skills`,
-  then plugin skill roots (earlier wins). `internal/skills` owns that merge via
-  `LoadFromRoots`; single-root `Load` remains for install/write.
-- **Hooks** can observe or block tool lifecycle events.
-
-## End-to-End Data Flow
-
-A typical interactive edit request looks like this:
-
-```mermaid
-sequenceDiagram
-    participant U as User
-    participant T as TUI
-    participant A as Agent loop
-    participant P as Provider
-    participant R as Tool registry
-    participant S as Sandbox
-    participant FS as Workspace/session files
-
-    U->>T: Type prompt and press Enter
-    T->>A: agent.Run(prompt, provider, options)
-    A->>P: CompletionRequest(messages, tool definitions)
-    P-->>A: Stream text + tool call edit_file
-    A->>S: Evaluate edit_file args
-    S-->>A: allow/prompt/deny decision
-    A->>T: permission callback if prompt needed
-    T-->>A: user decision
-    A->>R: Run edit_file
-    R->>FS: Write workspace file
-    R-->>A: Tool result + changed files
-    A->>P: Next CompletionRequest with tool result
-    P-->>A: Final answer
-    A-->>T: Result + callbacks
-    T->>FS: Append session events
-    T-->>U: Render final transcript
-```
-
-The same request in `zero exec` follows the same provider/tool/sandbox path, but
-streaming events are written to stdout/stderr instead of rendered in Bubble Tea.
-
-## Mental Model
-
-When you work on Zero, start by locating the change on the four-layer spine from
-[Architecture Spine](#architecture-spine) — Surface, Agent loop, Capabilities,
-Persistence — then jump to the package that owns that boundary:
-
-- UI behavior? Start in `internal/tui`.
-- Command-line behavior? Start in `internal/cli`.
-- Model/tool-turn behavior? Start in `internal/agent`.
-- A new capability? Register a tool in `internal/tools` or an extension surface.
-- Safety policy? Start in `internal/sandbox` and agent permission wiring.
-- Resume/history behavior? Start in `internal/sessions` and session recorders.

docs/INSTALL.md

@@ -4,11 +4,10 @@ Zero is distributed as:
 
 - an npm package, `@gitlawb/zero`
 - release archives on GitHub Releases
-- source builds with Go 1.26.5+
+- source builds with Go 1.25+
 
-The install scripts download a platform-specific release archive and require a
-published GitHub Release for the requested version. The npm package is
-self-contained: the platform binary installs from the npm registry.
+The npm package and install scripts download a platform-specific release archive.
+They require a published GitHub Release for the requested version.
 
 ## npm
 
@@ -17,32 +16,59 @@ npm install -g @gitlawb/zero
 zero
 ```
 
-The package supports Linux and macOS on x64 and arm64, and Windows on x64
-(Windows on ARM runs the x64 build under emulation). It installs a small
-`zero` wrapper plus, as an optional dependency, a platform payload with the
-native binary and the bundled browser/terminal control helpers. There are no
-install scripts and nothing is downloaded from outside the npm registry, so
-the install is silent and works identically under npm, Bun, pnpm, and yarn —
-no trust or approval steps. See [NPM_PACKAGING.md](NPM_PACKAGING.md) for the
-package architecture.
+The package supports Linux, macOS, and Windows on x64 and arm64. It installs the
+`zero` command and downloads the matching release binary during `postinstall`.
 
 Requirements:
 
 - Node.js 18+
-- network access to npm
+- network access to npm and GitHub Releases
 
-If the install skipped optional dependencies (`npm install --omit=optional`,
-or a package manager configured to do so), the wrapper falls back to
-downloading the binary from the matching GitHub Release, with checksum
-verification. The fetch is retried on each run until a binary is in place (a
-transient network failure heals itself; an install directory the current user
-cannot write to keeps failing — rerun the install or the command below with
-sufficient permissions). To trigger the fetch manually:
+## Bun
+
+Bun is "default-secure" and does not run lifecycle scripts of installed
+dependencies (only the installing project's own scripts), so the `postinstall`
+that fetches the Zero binary is silently skipped. The first run then fails with
+`No native binary found next to the npm wrapper`.
+
+The simplest fix is to trust the package after installing, which runs the
+blocked postinstall. This works for project and global installs:
 
 ```bash
-node "$(npm root -g)/@gitlawb/zero/scripts/postinstall.mjs"
+# project install
+bun add @gitlawb/zero
+bun pm trust @gitlawb/zero
+
+# global install
+bun add -g @gitlawb/zero
+bun pm -g trust @gitlawb/zero
 ```
 
+`bun pm untrusted` (or `bun pm -g untrusted`) lists the blocked postinstalls if
+you want to inspect before trusting.
+
+Alternatively, allow the postinstall to run at install time by adding the
+package to your project's `trustedDependencies` before installing:
+
+```json
+{
+  "trustedDependencies": ["@gitlawb/zero"]
+}
+```
+
+```bash
+bun add @gitlawb/zero
+```
+
+On Bun versions that do not have `bun pm trust`, run the installer manually
+after installing:
+
+```bash
+node node_modules/@gitlawb/zero/scripts/postinstall.mjs
+```
+
+Reference: <https://bun.sh/docs/pm/lifecycle>
+
 ## Linux And macOS Script
 
 Install the latest release:
@@ -125,7 +151,7 @@ Build a local binary:
 go build -o zero ./cmd/zero
 ```
 
-Source builds require Go 1.26.5+.
+Source builds require Go 1.25+.
 
 ### Sandbox Helpers For Source Builds
 
@@ -226,7 +252,8 @@ Supported targets:
 - `linux-arm64`
 - `macos-x64`
 - `macos-arm64`
-- `windows-x64` (Windows on ARM runs this build under emulation)
+- `windows-x64`
+- `windows-arm64`
 
 Each archive must have a matching `.sha256` file. The install scripts download
 both files, verify the checksum, and then copy the binary into the install
@@ -234,11 +261,10 @@ directory.
 
 ## Updating
 
-Check for a newer release, then download, verify, and install it in place:
+Check for a newer release:
 
 ```bash
 zero update --check
-zero upgrade
 ```
 
-See the [update guide](UPDATE.md) for update modes, flags, and platform details.
+Then reinstall with npm or rerun the install script for the version you want.

docs/NPM_PACKAGING.md

@@ -1,162 +0,0 @@
-# npm packaging
-
-How `@gitlawb/zero` is put together on npm, why it is shaped this way, and the
-rules the release pipeline must follow. Read this before touching
-`package.json`, `bin/zero.js`, `scripts/postinstall.mjs`,
-`scripts/npm/build-platform-packages.mjs`, or the npm-publish steps of
-`release-artifacts.yml`.
-
-## Goals
-
-A `npm install -g @gitlawb/zero` must be **silent and self-contained**:
-
-- No `EBADENGINE` warnings — ours or from any package in the dependency tree.
-- No install scripts anywhere in the tree, so npm's `allow-scripts` gating,
-  Bun's blocked-by-default lifecycle scripts, and pnpm's strict mode all
-  install zero without prompts, trust ceremonies, or broken binaries.
-- No network fetches outside the npm registry at install time. GitHub being
-  down or rate-limited must not break `npm install`.
-- Browser control (`agent-browser`) and terminal control (`tuistory`) work out
-  of the box.
-
-## Architecture
-
-The model is the one used by Codex, esbuild, and Biome: a tiny wrapper package
-plus per-platform payloads carrying the native binaries.
-
-```
-@gitlawb/zero                    <- wrapper: bin/zero.js + optionalDependencies
-├─ @gitlawb/zero-darwin-arm64 -> npm:@gitlawb/zero@{version}-darwin-arm64
-├─ @gitlawb/zero-darwin-x64   -> npm:@gitlawb/zero@{version}-darwin-x64
-├─ @gitlawb/zero-linux-arm64  -> npm:@gitlawb/zero@{version}-linux-arm64
-├─ @gitlawb/zero-linux-x64    -> npm:@gitlawb/zero@{version}-linux-x64
-└─ @gitlawb/zero-win32-x64    -> npm:@gitlawb/zero@{version}-win32-x64
-```
-
-- The platform "packages" are **versions of the same `@gitlawb/zero` package**,
-  published at suffixed versions (`0.4.0-linux-x64`) and referenced through
-  `npm:` aliases in `optionalDependencies`. One package name means one npm
-  trusted-publisher configuration — no new publish credentials per platform.
-  The alias suffixes use Node's `process.platform`/`process.arch` names so
-  `bin/zero.js` can derive its platform package directly.
-- Each platform version sets `os` and `cpu`, so npm installs exactly one of
-  them and skips the rest.
-- A platform payload is assembled from the platform's release archive by
-  `scripts/npm/build-platform-packages.mjs` and contains the `zero` binary,
-  the platform's sandbox helpers, and the vendored `helpers/` tree (see
-  below). It has no `bin`, no scripts, and no dependencies — it is inert on
-  its own; the wrapper execs the binary out of it.
-- The published wrapper has **no scripts and no dependencies** — the
-  assembly script strips both from the repo `package.json` and injects the
-  exact-version `optionalDependencies`. (The repo `package.json` keeps
-  `agent-browser`/`tuistory` in `dependencies` purely as the version pins for
-  the vendored helpers tree; they are never installed by consumers.)
-- There is **no windows-arm64 build** (matches the release matrix); Windows on
-  ARM runs the x64 build under emulation via the first-run fallback.
-
-### Binary resolution in `bin/zero.js`
-
-1. Resolve `@gitlawb/zero-<platform>-<arch>` and exec the `zero` binary from
-   it. This wins over any previously downloaded copy — the platform version is
-   pinned to the wrapper release.
-2. Fall back to a binary previously downloaded next to the wrapper.
-3. If neither exists (`--omit=optional`, package managers that skip optional
-   dependencies), run the **fallback downloader**: `scripts/postinstall.mjs`,
-   the exact logic that used to run as a postinstall hook (HTTPS-only, SHA-256
-   verified against the release's own checksum file, no zip-slip), invoked by
-   the wrapper itself. Failures are deliberately not cached — the fetch
-   retries on every run until a binary is in place, so a transient network
-   error self-heals. Platforms with no release asset (anything outside the
-   matrix above, including windows-arm64) skip the attempt entirely rather
-   than probing the network each run.
-4. If the download is impossible too, print build-from-source guidance.
-
-There is deliberately **no `scripts.postinstall`** in any published
-`package.json`. The downloader exists only as a first-run fallback.
-
-### Vendored helpers (`helpers/`)
-
-`agent-browser` (Apache-2.0, vercel-labs/agent-browser) and `tuistory` are
-**vendored binaries/packages inside the platform payload**, not npm
-dependencies of the wrapper:
-
-- As a dependency, agent-browser's `engines: { node: ">=24", pnpm: ">=11" }`
-  and postinstall script produce `EBADENGINE` and `allow-scripts` warnings for
-  every installer. As a vendored tree it produces none, because npm never
-  resolves it.
-- `zero-release package` already stages the helpers tree into every release
-  archive (`stageLocalControlHelpers` runs `npm ci` from the repo's
-  `package.json` pins + lockfile on the native builder). The assembly script
-  reuses that staged tree, so npm installs and `install.sh` installs get
-  identical helpers.
-- The Go binary resolves helpers from `<binary dir>/helpers/node_modules/.bin`
-  on its own (`internal/localcontrol/browser.go`, `adjacentHelper`) — no
-  wrapper involvement, no configuration.
-- **Symlink materialization:** `npm pack` silently drops symlinks, and npm's
-  own `.bin` shims are symlinks on POSIX. The assembly script rewrites every
-  `.bin` symlink into a relocatable `#!/bin/sh` exec shim and dereferences any
-  other symlink, so the vendored tree survives publishing. (Verified
-  empirically; nested `node_modules` under `helpers/` IS packed — only the
-  package-root `node_modules` is always ignored.)
-- **Binary pruning:** agent-browser ships one ~11 MB native binary per
-  platform (7 total). The assembly script keeps only the payload's own
-  platform binary (plus the musl variant on linux, which its launcher detects
-  at runtime) — about 65 MB saved per platform package.
-
-## Publishing rules (release pipeline)
-
-These are the invariants `release-artifacts.yml` must hold. Breaking the first
-one is user-visible immediately.
-
-1. **Platform versions must never become `latest`.** `0.4.0-linux-x64` is a
-   semver *prerelease* of `0.4.0`; publishing it without an explicit
-   non-`latest` dist-tag would clobber `latest` and users would install a
-   platform payload as the CLI. The workflow publishes platform versions with
-   `--tag platform` and asserts `latest` survived before the wrapper publish,
-   then asserts `latest` equals the wrapper version after it. Both lookups
-   tolerate a failed `npm view` (registry lag, or no `latest` tag yet on a
-   bootstrap publish) — a missing value must not abort the job between the
-   platform and wrapper publishes, which would strand a half-published
-   release.
-2. **Platform versions publish before the wrapper.** The wrapper's
-   `optionalDependencies` pin exact suffixed versions; publishing the wrapper
-   first would create a window where installs resolve aliases that 404.
-3. **Exact-version pinning.** The wrapper at `X.Y.Z` references platform
-   versions `X.Y.Z-<platform>-<arch>` exactly — never ranges — so a wrapper
-   and its binaries can never skew.
-4. All publishes go through the existing npm OIDC trusted-publishing flow and
-   the `npm-publish` environment gate; no tokens. Because the platform
-   payloads are versions of the same package, they reuse the single
-   trusted-publisher configuration.
-
-`zero update` keeps working unchanged: it detects an npm install by finding a
-`package.json` named `@gitlawb/zero` next to the running binary — true inside
-a platform payload too — and updates via `npm install -g @gitlawb/zero@latest`.
-
-## Runbook: bumping the vendored helpers
-
-The vendored helper versions are pinned by the repo `package.json`
-`dependencies` + `package-lock.json`; the release build vendors exactly what
-the lockfile resolves. No `^range` surprises ship: a release carries what its
-lockfile pinned at build time.
-
-- Routine bump: `npm install agent-browser@<version>` (or `tuistory@…`) in the
-  repo to update pin + lockfile, verify the browser/terminal tools end-to-end,
-  ship with the next release.
-- Security response: if a vendored helper publishes a security fix, the bump
-  is release-worthy on its own — cut a patch release of zero carrying only the
-  pin change.
-- Attribution: agent-browser is Apache-2.0; its LICENSE ships inside the
-  vendored package directory (`helpers/node_modules/agent-browser/`), which
-  satisfies redistribution attribution.
-
-## History / rationale
-
-Until v0.3.x the npm package was a wrapper with a `postinstall` downloader and
-`agent-browser`/`tuistory` as regular dependencies. That produced three
-warnings on every install (`EBADENGINE` from agent-browser's `node >=24`
-engines pin, and `allow-scripts` warnings for both postinstall scripts), broke
-silently under Bun's default script blocking, and coupled `npm install` to
-GitHub Releases availability. The platform-package model removes the warnings
-structurally — there is nothing left in the tree for a package manager to warn
-about — rather than asking users to approve or suppress them.

docs/NPM_WRAPPER_SMOKE.md

@@ -1,8 +1,8 @@
 # npm Wrapper Smoke Checklist
 
 Run this checklist when a PR changes npm distribution files such as
-`package.json`, `bin/zero.js`, `scripts/npm/build-platform-packages.mjs`,
-Go build or release tooling, or the npm `bin` wrapper.
+`package.json`, `bin/zero.js`, Go build or release tooling, or the npm `bin`
+wrapper.
 
 ## Required Checks
 
@@ -21,52 +21,17 @@ go run ./cmd/zero-release build
 go run ./cmd/zero-release smoke
 ```
 
-## End-To-End Assembly Smoke
-
-Build a real archive for the host platform and assemble the npm payloads from
-it (this is exactly what the publish job does):
-
-```bash
-go run ./cmd/zero-release package
-node scripts/npm/build-platform-packages.mjs \
-  --artifacts-dir dist/release --out-dir dist/npm --only <platform>-<arch>
-```
-
-Then simulate an installed layout and run the wrapper through each resolution
-path:
-
-```bash
-# platform-package path
-mkdir -p /tmp/zero-sim/node_modules/@gitlawb
-cp -R dist/npm/wrapper /tmp/zero-sim/node_modules/@gitlawb/zero
-cp -R dist/npm/platforms/zero-<platform>-<arch> \
-  "/tmp/zero-sim/node_modules/@gitlawb/zero-<platform>-<arch>"
-node /tmp/zero-sim/node_modules/@gitlawb/zero/bin/zero.js --version
-
-# first-run download fallback (delete the platform package first)
-rm -rf "/tmp/zero-sim/node_modules/@gitlawb/zero-<platform>-<arch>"
-node /tmp/zero-sim/node_modules/@gitlawb/zero/bin/zero.js --version
-```
-
 ## Checklist
 
 - `package.json` has the expected package name (`@gitlawb/zero`), version,
-  `bin.zero` entry, and NO `scripts` entries — the published package must be
-  free of lifecycle scripts (see NPM_PACKAGING.md).
-- `scripts/npm/build-platform-packages.mjs` emits platform payloads whose
-  `package.json` is `@gitlawb/zero@<version>-<platform>-<arch>` with matching
-  `os`/`cpu`, contains the executable binary and the vendored `helpers/` tree
-  with regular-file (non-symlink) `.bin` shims, and a wrapper payload with no
-  scripts, no dependencies, and the full five-alias `optionalDependencies`
-  matrix.
-- `scripts/postinstall.mjs` (the first-run fallback) resolves the correct
-  release asset name/URL per platform (`ZERO_INSTALL_DRY_RUN=1` prints the
-  plan), verifies the downloaded archive's SHA-256, and extracts only the
-  known binary basenames into place. `ZERO_SKIP_DOWNLOAD=1` opts out cleanly
-  (exit 0) and an unsupported platform/arch is a non-fatal skip.
-- The wrapper prefers the platform package binary over a previously downloaded
-  one, and `agent-browser`/`tuistory` run from the platform package's
-  `helpers/node_modules/.bin` shims.
+  `bin.zero` entry, and exactly one `scripts` entry — the `postinstall` hook.
+- `scripts/postinstall.mjs` resolves the correct release asset name/URL per
+  platform (`ZERO_INSTALL_DRY_RUN=1` prints the plan), verifies the downloaded
+  archive's SHA-256, and extracts only the known binary basenames into place.
+  `ZERO_SKIP_DOWNLOAD=1` opts out cleanly (exit 0) and an unsupported
+  platform/arch is a non-fatal skip.
+- The wrapper binary resolves through the package `bin` entry and
+  `node_modules/.bin` in a package-install smoke test.
 - The built binary exits 0 for `zero --version` or `zero --help`.
 - `zero --version` reports `zero <package.json version>`.
 - Release packaging still emits the expected archive and checksum names when

docs/README.md

@@ -17,7 +17,6 @@ Zero.
 
 ## Maintainer Docs
 
-- [How Zero Works](HOW_ZERO_WORKS.md)
 - [npm wrapper smoke checklist](NPM_WRAPPER_SMOKE.md)
 - [Benchmarks](BENCHMARK.md)
 - [Performance benchmarks](PERFORMANCE.md)

docs/SPECIALISTS.md

@@ -138,45 +138,3 @@ output or stop a still-running task by id.
 If Zero is restarted while a background task is still marked `running`, the new
 manager marks that task `error` and clears its PID. This avoids sending
 `TaskStop` to a stale PID that may now belong to an unrelated process.
-
-## Recovering an Interrupted Overwrite
-
-Zero writes and flushes a complete temporary file before publishing an overwrite,
-so a write failure before publication leaves the existing manifest unchanged
-instead of truncating it. On Unix, publication uses a same-directory rename and
-preserves the existing file's permission bits.
-
-On Windows, Zero uses `ReplaceFileW` to preserve the destination DACL instead of
-silently replacing it with the temporary file's inherited DACL. `ReplaceFileW`
-is not observer-atomic: another process can briefly observe the destination path
-as absent during replacement. Zero serializes specialist loads and managed
-mutations within one process, but cannot synchronize external processes or
-editors.
-
-Windows errors 1176 (`ERROR_UNABLE_TO_MOVE_REPLACEMENT`) and 1177
-(`ERROR_UNABLE_TO_MOVE_REPLACEMENT_2`) are partial replacement failures. With
-Zero's managed backup, 1176 leaves the original names intact and needs no manual
-recovery. For 1177, Zero has moved the original aside and tries to move it back.
-That rollback almost always succeeds, and the failed write changes nothing.
-
-If the rollback itself fails — typically because another process is holding a
-lock on the file — the original is not lost, but it is left under a name Zero
-does not read:
-
-```text
-<specialist dir>/.zero-replace-<random>.backup
-```
-
-Only `*.md` files are loaded as specialists, so until that file is renamed the
-specialist will not appear in `zero specialist list` or resolve by name. Zero's
-error message includes both the backup path and destination path. Recover by
-closing whatever holds the lock and renaming the backup back:
-
-```powershell
-Move-Item .zero-replace-<random>.backup <name>.md
-```
-
-A `.zero-replace-*.backup` can also linger after a *successful* overwrite if the
-backup could not be deleted afterward. That case is reported as a warning rather
-than an error — the new manifest is already in place, and the leftover file is
-safe to delete.

docs/THEMES.md

@@ -1,39 +0,0 @@
-# TUI Themes
-
-Zero's TUI ships a set of built-in color themes. Pick one with `--theme <name>`,
-the `ZERO_THEME` environment variable, or the `/theme <name>` command while
-running. `auto` (the default) follows the terminal's detected background.
-
-Run `/theme` with no argument to open a picker: move through the list to
-live-preview each theme, press Enter to apply the highlighted one, or Esc to
-cancel and restore the previously active theme. Run `/theme list` to print
-the active theme and the registered names without opening the picker.
-
-## Dune (`dune`)
-
-A warm sand-and-cream palette: sand/cream surface, charcoal ink, and a soft
-amber accent.
-
-## Neon (`neon`)
-
-A neon-on-black palette: pitch-black surface, bright green ink, and a cyan accent.
-
-## Adding a theme
-
-Every theme is a `palette` (a table of color hex tokens) plus one entry in
-`themeRegistry`, both in `internal/tui/theme_palettes.go`. `buildTheme` in
-`internal/tui/theme.go` turns a `palette` into the resolved `lipgloss.Style`
-set every renderer reads from the active `zeroTheme`. Adding a theme means
-adding a new `palette{...}` literal, a `themeRegistry` entry, and test
-coverage for the new palette (see below).
-
-Registry-wide tests in `internal/tui/theme_select_test.go` assert the basic
-WCAG AA text tokens, the gray-ramp order, the diff word-span pairs, and the
-selected-row band for every entry. The rendered-surface invariants beyond
-those (permission surfaces, selected-row secondary text, diff gutters, and
-the xterm-256 downsampling checks) are asserted per palette, not against the
-whole registry: `TestExtendedThemeContrastInvariants` and
-`TestExtendedThemeANSI256Contrast` enumerate the palettes they cover. A new
-theme must be added to those tests (or given equivalent palette-specific
-assertions), or CI can stay green while its permission, selected-row, and
-diff surfaces ship unreadable.

docs/UPDATE.md

@@ -1,23 +1,16 @@
 # Update Flow
 
 `zero update --check` checks the latest GitHub release and compares it with the
-local CLI version. `zero update --apply` (or its shorthand, `zero upgrade`)
-downloads, verifies, and installs it.
+local CLI version.
 
 ```bash
 zero update --check
 zero update --check --json
 zero update --check --repo Gitlawb/zero
 zero update --check --target windows-x64
-
-zero upgrade
-zero update --apply
 ```
 
-`--check` and `--apply` are mutually exclusive. `zero update` requires one of
-them explicitly; `zero upgrade` is `zero update` with `--apply` implied.
-
-`--check` is check-only:
+The command is intentionally check-only:
 
 - It does not replace the running binary.
 - It exits with code `0` when the check succeeds, even when an update is
@@ -26,35 +19,14 @@ them explicitly; `zero upgrade` is `zero update` with `--apply` implied.
 - `--json` prints the same result in a machine-readable format for scripts and
   CI.
 
-`--apply` installs the update in place:
-
-- npm installs delegate to `npm install -g @gitlawb/zero@latest`.
-- Standalone installs download the release archive, verify its checksum,
-  extract it, and atomically replace the running binary plus any installed
-  optional sandbox helpers.
-- On Windows, the running executable is renamed aside and cleaned up on the
-  next `zero update --apply` or `zero upgrade`, since it can't be overwritten
-  while running.
-- `--target` cannot be combined with `--apply`; it only applies to `--check`,
-  since applying always installs onto the current machine.
-- `--repo` and `--endpoint` are ignored when applying to an npm-managed
-  install: that path delegates to `npm install -g @gitlawb/zero@latest` and
-  takes its release from the npm registry, not from GitHub. They still apply to
-  `--check` there.
-- `--json` serializes Zero's final result. For npm-managed installs, npm may
-  also write progress output to stdout, so neither
-  `zero update --apply --json` nor `zero upgrade --json` is guaranteed to
-  produce a single parseable JSON document. Use `--check --json` for
-  machine-readable automation.
-
 Useful flags:
 
 | Flag | Purpose |
 |---|---|
-| `--repo <owner/repo>` | Use another GitHub repository for `--check` and `--apply`/`upgrade`. Ignored by an npm-managed apply. |
-| `--endpoint <url\|owner/repo>` | Use a specific release API URL or repository slug for `--check` and `--apply`/`upgrade`. Ignored by an npm-managed apply. |
+| `--repo <owner/repo>` | Check another GitHub repository. |
+| `--endpoint <url|owner/repo>` | Check a specific release API URL or repository slug. |
 | `--timeout <duration>` | Override the default release check timeout. |
-| `--target <platform-arch>` | Validate release metadata for another supported target (`--check` only). |
+| `--target <platform-arch>` | Validate release metadata for another supported target. |
 
 Supported targets are `linux-x64`, `linux-arm64`, `macos-x64`, `macos-arm64`,
 `windows-x64`, and `windows-arm64`. Without `--target`, Zero checks the current
@@ -68,45 +40,5 @@ Endpoint resolution order:
 4. `https://api.github.com/repos/Gitlawb/zero/releases/latest`
 
 Installer scripts download the matching release asset for the local platform and
-verify its `.sha256` file. If Zero is already installed, run `zero upgrade`
-instead of reinstalling.
-
-## Windows recovery state (standalone installs)
-
-This section describes Windows only. On Linux and macOS a standalone update
-renames the staged file directly over the executable path through the
-installation directory's file descriptor, so the replacement is atomic and no
-aside copy, marker, or recovery record is ever created — a failed update leaves
-the previous binary in place and nothing to resolve.
-
-On Windows, a running executable cannot be replaced in place, so the previous
-binary is moved aside to `<binary>.zero-update-<random>.old` first. The updater
-records that exact file — bound to its filesystem identity, in per-user state
-outside the installation directory — and deletes only that recorded copy after
-the next update is verified in place. Backups it did not create are never
-removed, so a file such as `zero.exe.before-manual-patch.old` is left alone. A
-recorded copy that something else holds open (an on-access scanner, an editor)
-stays recorded and is removed by a later update instead.
-
-An update **refuses to run** while unresolved recovery state exists beside the
-binary:
-
-| State on disk | Meaning |
-|---|---|
-| `<binary>.old` (or `<binary>.<suffix>.old`) plus a `.keep` marker | A previous update could not restore the original binary. The `.old` file may be the last binary the updater verified; the installed one may be unverified. |
-| `<binary>.…old.<suffix>.recovery` | A previous update could not even write the marker, so it moved the last verified binary to that name. |
-| The binary is missing and one or more `*.old` files exist | The previous attempt was interrupted between moves. |
-
-The refusal names the paths involved and the two moves that end the state:
-either move the recovery binary back over the executable path, or — if the
-installed binary is the one you want — delete the `.keep` marker (or the
-`.recovery` copy, once you have verified the installed binary) and update again.
-
-This is deliberately fail-closed and differs from older releases, which deleted
-`<binary>.old` and proceeded. The trade-off is that anyone who can write in the
-installation directory can plant `<binary>.old` and `<binary>.old.keep` there
-and make every subsequent update refuse until an operator clears them. That is
-the safe direction: the alternative is an update that overwrites the only
-verified copy of the previous binary. If updates start refusing, inspect those
-files before removing them — an installation directory that a lower-privileged
-account can write to is itself worth fixing.
+verify its `.sha256` file. If Zero is already installed, run `zero update --check`
+before reinstalling.

go.mod

@@ -1,33 +1,32 @@
 module github.com/Gitlawb/zero
 
-go 1.26.6
+go 1.25.0
+
+toolchain go1.26.4
 
 require (
-	charm.land/bubbles/v2 v2.1.1
-	charm.land/bubbletea/v2 v2.0.8
-	charm.land/lipgloss/v2 v2.0.5
-	github.com/Microsoft/go-winio v0.6.2
+	charm.land/bubbles/v2 v2.1.0
+	charm.land/bubbletea/v2 v2.0.7
+	charm.land/lipgloss/v2 v2.0.4
 	github.com/alecthomas/chroma/v2 v2.27.0
 	github.com/atotto/clipboard v0.1.4
 	github.com/aymanbagabas/go-udiff v0.4.1
 	github.com/charmbracelet/colorprofile v0.4.3
-	github.com/charmbracelet/ultraviolet v0.0.0-20260703014108-f5a850f9c2b7
 	github.com/charmbracelet/x/ansi v0.11.7
 	github.com/charmbracelet/x/term v0.2.2
-	github.com/coder/websocket v1.8.15
 	github.com/ledongthuc/pdf v0.0.0-20250511090121-5959a4027728
-	golang.org/x/image v0.45.0
-	golang.org/x/sys v0.47.0
+	golang.org/x/sys v0.46.0
 	mvdan.cc/sh/v3 v3.13.1
 )
 
 require (
+	github.com/charmbracelet/ultraviolet v0.0.0-20260615092913-2399af76d5b1 // indirect
 	github.com/charmbracelet/x/exp/golden v0.0.0-20260615092313-b57e5e6d29bb // indirect
 	github.com/charmbracelet/x/termios v0.1.1 // indirect
 	github.com/charmbracelet/x/windows v0.2.2 // indirect
 	github.com/clipperhouse/displaywidth v0.11.0 // indirect
 	github.com/clipperhouse/uax29/v2 v2.7.0 // indirect
-	github.com/dlclark/regexp2/v2 v2.5.0 // indirect
+	github.com/dlclark/regexp2/v2 v2.2.1 // indirect
 	github.com/go-quicktest/qt v1.102.0 // indirect
 	github.com/lucasb-eyer/go-colorful v1.4.0 // indirect
 	github.com/mattn/go-runewidth v0.0.24 // indirect
@@ -36,5 +35,5 @@ require (
 	github.com/rogpeppe/go-internal v1.15.0 // indirect
 	github.com/xo/terminfo v0.0.0-20220910002029-abceb7e1c41e // indirect
 	golang.org/x/exp v0.0.0-20260611194520-c48552f49976 // indirect
-	golang.org/x/sync v0.22.0 // indirect
+	golang.org/x/sync v0.21.0 // indirect
 )

go.sum

@@ -1,11 +1,9 @@
-charm.land/bubbles/v2 v2.1.1 h1:7r55WzBxpo/R3z98hGmY7KKPd3ET6vsf0Fb9sDHOV60=
-charm.land/bubbles/v2 v2.1.1/go.mod h1:GE6M31gaWZVXzGw73OeuTTgy4lX+OtkH0E5ymnNsHxo=
-charm.land/bubbletea/v2 v2.0.8 h1:SxTJMhCAI3lbPmy4SgX5LWZ24AdINr4I6UEqzZvYJuY=
-charm.land/bubbletea/v2 v2.0.8/go.mod h1:2SkdgoTXluXJHOUwAoRlRXF/28vklb1rFl6GcgV1/ss=
-charm.land/lipgloss/v2 v2.0.5 h1:kbNxgeeUOYv5J0YdpxFjfvf3dFvqH8Aci4zB6xqFtrY=
-charm.land/lipgloss/v2 v2.0.5/go.mod h1:9oqhxt4yxIMe6q5A4kHr44DremZk7J9UNh74GlWa5nc=
-github.com/Microsoft/go-winio v0.6.2 h1:F2VQgta7ecxGYO8k3ZZz3RS8fVIXVxONVUPlNERoyfY=
-github.com/Microsoft/go-winio v0.6.2/go.mod h1:yd8OoFMLzJbo9gZq8j5qaps8bJ9aShtEA8Ipt1oGCvU=
+charm.land/bubbles/v2 v2.1.0 h1:YSnNh5cPYlYjPxRrzs5VEn3vwhtEn3jVGRBT3M7/I0g=
+charm.land/bubbles/v2 v2.1.0/go.mod h1:l97h4hym2hvWBVfmJDtrEHHCtkIKeTEb3TTJ4ZOB3wY=
+charm.land/bubbletea/v2 v2.0.7 h1:7qw2tTAVar7m7klOPBYfTB0mniv/RuexsYwMRNxSeL0=
+charm.land/bubbletea/v2 v2.0.7/go.mod h1:DGW2q8gvzHnOpMpZTORs0aySVHCox5C+2Svk0fci1qs=
+charm.land/lipgloss/v2 v2.0.4 h1:lcPeVtcp23SNra7lHy8iYE4UC2aIipVQ47sbGyyxR5Q=
+charm.land/lipgloss/v2 v2.0.4/go.mod h1:0653x8epbZSzdDfO/XPS1a/uYPOBeSsCssOpJOqDzik=
 github.com/alecthomas/assert/v2 v2.11.0 h1:2Q9r3ki8+JYXvGsDyBXwH3LcJ+WK5D0gc5E8vS6K3D0=
 github.com/alecthomas/assert/v2 v2.11.0/go.mod h1:Bze95FyfUr7x34QZrjL+XP+0qgp/zg8yS+TtBj1WA3k=
 github.com/alecthomas/chroma/v2 v2.27.0 h1:FodwmyOBgJULFYmDqibcp9pvfDLWdtPRh9v/r5BXYZs=
@@ -18,8 +16,8 @@ github.com/aymanbagabas/go-udiff v0.4.1 h1:OEIrQ8maEeDBXQDoGCbbTTXYJMYRCRO1fnodZ
 github.com/aymanbagabas/go-udiff v0.4.1/go.mod h1:0L9PGwj20lrtmEMeyw4WKJ/TMyDtvAoK9bf2u/mNo3w=
 github.com/charmbracelet/colorprofile v0.4.3 h1:QPa1IWkYI+AOB+fE+mg/5/4HRMZcaXex9t5KX76i20Q=
 github.com/charmbracelet/colorprofile v0.4.3/go.mod h1:/zT4BhpD5aGFpqQQqw7a+VtHCzu+zrQtt1zhMt9mR4Q=
-github.com/charmbracelet/ultraviolet v0.0.0-20260703014108-f5a850f9c2b7 h1:3FmWoGNWK4STvqg0O0Aeav2T7rodWJAPeF0QpH+8gFw=
-github.com/charmbracelet/ultraviolet v0.0.0-20260703014108-f5a850f9c2b7/go.mod h1:f/jRa757WUmaOZrbPspXymbg/GnbF+rwe4OLsG7aXYo=
+github.com/charmbracelet/ultraviolet v0.0.0-20260615092913-2399af76d5b1 h1:4+r3uOJ69ueRBt4okgEfWZeXs3BD36HcDBmOIAUlETk=
+github.com/charmbracelet/ultraviolet v0.0.0-20260615092913-2399af76d5b1/go.mod h1:f/jRa757WUmaOZrbPspXymbg/GnbF+rwe4OLsG7aXYo=
 github.com/charmbracelet/x/ansi v0.11.7 h1:kzv1kJvjg2S3r9KHo8hDdHFQLEqn4RBCb39dAYC84jI=
 github.com/charmbracelet/x/ansi v0.11.7/go.mod h1:9qGpnAVYz+8ACONkZBUWPtL7lulP9No6p1epAihUZwQ=
 github.com/charmbracelet/x/exp/golden v0.0.0-20260615092313-b57e5e6d29bb h1:m1Uub9uM7meZ8ofVyeRfK7M/knCi3sF2gKn9kldVW0I=
@@ -34,10 +32,8 @@ github.com/clipperhouse/displaywidth v0.11.0 h1:lBc6kY44VFw+TDx4I8opi/EtL9m20WSE
 github.com/clipperhouse/displaywidth v0.11.0/go.mod h1:bkrFNkf81G8HyVqmKGxsPufD3JhNl3dSqnGhOoSD/o0=
 github.com/clipperhouse/uax29/v2 v2.7.0 h1:+gs4oBZ2gPfVrKPthwbMzWZDaAFPGYK72F0NJv2v7Vk=
 github.com/clipperhouse/uax29/v2 v2.7.0/go.mod h1:EFJ2TJMRUaplDxHKj1qAEhCtQPW2tJSwu5BF98AuoVM=
-github.com/coder/websocket v1.8.15 h1:6B2JPeOGlpff2Uz6vOEH1Vzpi0iUz20A+lPVhPHtNUA=
-github.com/coder/websocket v1.8.15/go.mod h1:NX3SzP+inril6yawo5CQXx8+fk145lPDC6pumgx0mVg=
-github.com/dlclark/regexp2/v2 v2.5.0 h1:liNiWIPCvCh5HBcYfsjd+P16AG79fwd6T1Toy2gOtEA=
-github.com/dlclark/regexp2/v2 v2.5.0/go.mod h1:avUrQvPaLz2DrFNHJF0taWAFFX2C1GMSSoeiqFjcBmU=
+github.com/dlclark/regexp2/v2 v2.2.1 h1:mf4KkFUj0gJuarK8P+LgiS+Lit7m9N1yAwEfPbee7R0=
+github.com/dlclark/regexp2/v2 v2.2.1/go.mod h1:avUrQvPaLz2DrFNHJF0taWAFFX2C1GMSSoeiqFjcBmU=
 github.com/go-quicktest/qt v1.102.0 h1:HSQxCeh5YZH3EL3W39ixjtyaEhcWSXQHtHnMBzSs474=
 github.com/go-quicktest/qt v1.102.0/go.mod h1:p4lGIVX+8Wa6ZPNDvqcxq36XpUDLh42FLetFU7odllI=
 github.com/google/go-cmp v0.7.0 h1:wk8382ETsv4JYUZwIsn6YpYiWiBsYLSJiTsyBybVuN8=
@@ -64,13 +60,9 @@ github.com/xo/terminfo v0.0.0-20220910002029-abceb7e1c41e h1:JVG44RsyaB9T2KIHavM
 github.com/xo/terminfo v0.0.0-20220910002029-abceb7e1c41e/go.mod h1:RbqR21r5mrJuqunuUZ/Dhy/avygyECGrLceyNeo4LiM=
 golang.org/x/exp v0.0.0-20260611194520-c48552f49976 h1:X8Hz2ImujgbmetVuW+w2YkyZChE3cBpZi2P158rTG9M=
 golang.org/x/exp v0.0.0-20260611194520-c48552f49976/go.mod h1:vnf4pv9iKZXY58sQE1L86zmNWJ4159e1RkcWiLCkeEY=
-golang.org/x/image v0.44.0 h1:+tDekMZED9+LrtB3G5xzRggpVh9CARjZqROla3R3R+I=
-golang.org/x/image v0.44.0/go.mod h1:V8K3KE9KKKE+pLpQDOeN18w9oacNSvy1tDOirTu4xtY=
-golang.org/x/image v0.45.0 h1:FMb1nTbH5H9vF55SriQHgFw5GnNL9Jg6L25BwXKzhB0=
-golang.org/x/image v0.45.0/go.mod h1:n62x/7RqlwXDvGsSU4u6IUTUf6KghUZ9Bt7cG/T9Fx4=
-golang.org/x/sync v0.22.0 h1:SZjpbeLmrCk4xhRSZFNZW5gFUeCeFgjekvI/+gfScek=
-golang.org/x/sync v0.22.0/go.mod h1:9xrNwdLfx4jkKbNva9FpL6vEN7evnE43NNNJQ2LF3+0=
-golang.org/x/sys v0.47.0 h1:o7XGOvZQCADBQQ4Y7VNq2dRWQR7JmOUW8Kxx4ZsNgWs=
-golang.org/x/sys v0.47.0/go.mod h1:4GL1E5IUh+htKOUEOaiffhrAeqysfVGipDYzABqnCmw=
+golang.org/x/sync v0.21.0 h1:HLII4xRRTtCRkxYp4HNFF0Js/Og6q2i++KXbg0gHCwM=
+golang.org/x/sync v0.21.0/go.mod h1:9xrNwdLfx4jkKbNva9FpL6vEN7evnE43NNNJQ2LF3+0=
+golang.org/x/sys v0.46.0 h1:noSf2Fq6F8DBgS+LysIkx7rIExoNHJsxOAtPp4rthXw=
+golang.org/x/sys v0.46.0/go.mod h1:4GL1E5IUh+htKOUEOaiffhrAeqysfVGipDYzABqnCmw=
 mvdan.cc/sh/v3 v3.13.1 h1:DP3TfgZhDkT7lerUdnp6PTGKyxxzz6T+cOlY/xEvfWk=
 mvdan.cc/sh/v3 v3.13.1/go.mod h1:lXJ8SexMvEVcHCoDvAGLZgFJ9Wsm2sulmoNEXGhYZD0=

internal/acp/agent.go

@@ -5,15 +5,11 @@ import (
 	"encoding/base64"
 	"encoding/json"
 	"errors"
-	"log"
 	"strings"
 	"sync"
 
 	"github.com/Gitlawb/zero/internal/agent"
 	"github.com/Gitlawb/zero/internal/config"
-	"github.com/Gitlawb/zero/internal/providercatalog"
-	"github.com/Gitlawb/zero/internal/providermodelcatalog"
-	"github.com/Gitlawb/zero/internal/providermodeldiscovery"
 	"github.com/Gitlawb/zero/internal/sandbox"
 	"github.com/Gitlawb/zero/internal/sessions"
 	"github.com/Gitlawb/zero/internal/tools"
@@ -26,10 +22,9 @@ import (
 // means the editor only hosts the thread — ZERO owns BYOK and telemetry-free
 // operation.
 type Deps struct {
-	ResolveConfig  func(workspaceRoot string, overrides config.Overrides) (config.ResolvedConfig, error)
-	DiscoverModels func(context.Context, config.ProviderProfile) ([]providermodeldiscovery.Model, error)
-	NewProvider    func(profile config.ProviderProfile) (zeroruntime.Provider, error)
-	RunAgent       func(ctx context.Context, prompt string, provider zeroruntime.Provider, opts agent.Options) (agent.Result, error)
+	ResolveConfig func(workspaceRoot string, overrides config.Overrides) (config.ResolvedConfig, error)
+	NewProvider   func(profile config.ProviderProfile) (zeroruntime.Provider, error)
+	RunAgent      func(ctx context.Context, prompt string, provider zeroruntime.Provider, opts agent.Options) (agent.Result, error)
 	// BuildWorkspace builds the SCOPED tool registry and the sandbox engine for a
 	// validated workspace root, so ACP shell tools (bash/exec_command) are confined
 	// exactly like the exec surface — never run unconfined on the host.
@@ -64,16 +59,13 @@ type acpSession struct {
 	// turnMu serializes prompt turns for one session: concurrent session/prompt
 	// calls run one at a time so they can't interleave history or clobber the
 	// single cancel slot.
-	turnMu  sync.Mutex
-	modelMu sync.Mutex
+	turnMu sync.Mutex
 
-	mu             sync.Mutex
-	mode           agent.PermissionMode
-	model          string
-	models         []SessionConfigOptionValue
-	restrictModels bool
-	cancel         context.CancelFunc
-	history        []turnRecord
+	mu      sync.Mutex
+	mode    agent.PermissionMode
+	model   string // override; "" => config default
+	cancel  context.CancelFunc
+	history []turnRecord
 }
 
 // NewAgent builds the ACP server and registers its method handlers on conn.
@@ -90,6 +82,9 @@ func NewAgent(conn *Conn, deps Deps) *Agent {
 	return a
 }
 
+// Serve runs the connection read loop until the stream closes or ctx is done.
+func (a *Agent) Serve(ctx context.Context) error { return a.conn.Serve(ctx) }
+
 // ---- initialize ----
 
 func (a *Agent) handleInitialize(_ context.Context, params json.RawMessage) (any, error) {
@@ -124,7 +119,7 @@ func (a *Agent) handleInitialize(_ context.Context, params json.RawMessage) (any
 
 // ---- session lifecycle ----
 
-func (a *Agent) handleSessionNew(ctx context.Context, params json.RawMessage) (any, error) {
+func (a *Agent) handleSessionNew(_ context.Context, params json.RawMessage) (any, error) {
 	var p NewSessionParams
 	if err := json.Unmarshal(params, &p); err != nil {
 		return nil, RPCError(codeInvalidParams, "invalid session/new params")
@@ -133,23 +128,18 @@ func (a *Agent) handleSessionNew(ctx context.Context, params json.RawMessage) (a
 	if err != nil {
 		return nil, RPCError(codeInvalidParams, err.Error())
 	}
-	model, models, restrictModels, err := a.resolveModelChoices(ctx, root)
-	if err != nil {
-		return nil, RPCError(codeInternalError, "config: "+err.Error())
-	}
-	meta, err := a.deps.Store.Create(sessions.CreateInput{Title: "ACP session", Cwd: root, ModelID: model})
+	meta, err := a.deps.Store.Create(sessions.CreateInput{Title: "ACP session", Cwd: root})
 	if err != nil {
 		return nil, RPCError(codeInternalError, "create session: "+err.Error())
 	}
-	sess := a.registerSession(meta.SessionID, root, nil, model, models, restrictModels)
+	sess := a.registerSession(meta.SessionID, root, nil)
 	return NewSessionResult{
-		SessionID:     sess.id,
-		ConfigOptions: a.configOptions(sess),
-		Modes:         a.modeState(sess),
+		SessionID: sess.id,
+		Modes:     a.modeState(sess),
 	}, nil
 }
 
-func (a *Agent) handleSessionLoad(ctx context.Context, params json.RawMessage) (any, error) {
+func (a *Agent) handleSessionLoad(_ context.Context, params json.RawMessage) (any, error) {
 	var p LoadSessionParams
 	if err := json.Unmarshal(params, &p); err != nil {
 		return nil, RPCError(codeInvalidParams, "invalid session/load params")
@@ -169,27 +159,10 @@ func (a *Agent) handleSessionLoad(ctx context.Context, params json.RawMessage) (
 	// Load history BEFORE publishing the session so no concurrent prompt observes
 	// a half-initialized session (registerSession sets history under the lock and
 	// reuses an already-live session rather than orphaning its in-flight turn).
-	history, historyErr := a.loadHistory(meta.SessionID)
-	model, models, restrictModels, err := a.resolveModelChoices(ctx, root)
-	if err != nil {
-		return nil, RPCError(codeInternalError, "config: "+err.Error())
-	}
-	if persistedModel := strings.TrimSpace(meta.ModelID); persistedModel != "" && (!restrictModels || modelChoiceExists(models, persistedModel)) {
-		model = persistedModel
-		if !modelChoiceExists(models, persistedModel) {
-			models = append(models, SessionConfigOptionValue{Value: persistedModel, Name: persistedModel})
-		}
-	}
-	sess := a.registerSession(meta.SessionID, root, history, model, models, restrictModels)
-	a.warnPersistence(
-		&notifier{conn: a.conn, sessionID: sess.id},
-		"load session history",
-		"Could not load session history. The session is open, but earlier turns may be missing until storage recovers.",
-		historyErr,
-	)
+	history := a.loadHistory(meta.SessionID)
+	sess := a.registerSession(meta.SessionID, root, history)
 	return LoadSessionResult{
-		ConfigOptions: a.configOptions(sess),
-		Modes:         a.modeState(sess),
+		Modes: a.modeState(sess),
 	}, nil
 }
 
@@ -280,14 +253,7 @@ func (a *Agent) runTurn(ctx context.Context, sess *acpSession, userText string,
 	if stopErr != nil {
 		return "", RPCError(codeInternalError, stopErr.Error())
 	}
-	if err := a.persistTurn(sess, userText, result.FinalAnswer); err != nil {
-		a.warnPersistence(
-			note,
-			"save session history",
-			"Could not save session history. This turn is available in memory, but future resume may miss it until storage recovers.",
-			err,
-		)
-	}
+	a.persistTurn(sess, userText, result.FinalAnswer)
 	return reason, nil
 }
 
@@ -349,11 +315,9 @@ func (a *Agent) handleSetMode(_ context.Context, params json.RawMessage) (any, e
 	if sess == nil {
 		return nil, RPCError(codeInvalidParams, "unknown session: "+p.SessionID)
 	}
-	sess.turnMu.Lock()
-	defer sess.turnMu.Unlock()
 	mode := agent.PermissionMode(p.ModeID)
 	switch mode {
-	case agent.PermissionModeAuto, agent.PermissionModeAsk, agent.PermissionModePlan:
+	case agent.PermissionModeAuto, agent.PermissionModeAsk:
 		sess.setMode(mode)
 		(&notifier{conn: a.conn, sessionID: sess.id}).currentMode(string(mode))
 		return SetSessionModeResult{}, nil
@@ -376,31 +340,11 @@ func (a *Agent) handleSetConfigOption(_ context.Context, params json.RawMessage)
 	if sess == nil {
 		return nil, RPCError(codeInvalidParams, "unknown session: "+p.SessionID)
 	}
-	switch p.ConfigID {
-	case configIDModel:
-		model := strings.TrimSpace(p.Value)
-		if err := a.updateModel(sess, model, sess.restrictModels); err != nil {
-			return nil, err
-		}
-	case configIDMode:
-		// Same turnMu as handleSetMode so the two advertised mode doors (session
-		// set_mode and set_config_option) serialize mode flips consistently.
-		sess.turnMu.Lock()
-		defer sess.turnMu.Unlock()
-		mode := agent.PermissionMode(p.Value)
-		switch mode {
-		case agent.PermissionModeAuto, agent.PermissionModeAsk, agent.PermissionModePlan:
-			sess.setMode(mode)
-			(&notifier{conn: a.conn, sessionID: sess.id}).currentMode(string(mode))
-		case agent.PermissionModeUnsafe:
-			return nil, RPCError(codeInvalidParams, "mode not permitted over ACP: "+p.Value)
-		default:
-			return nil, RPCError(codeInvalidParams, "unknown mode: "+p.Value)
-		}
-	default:
+	if p.ConfigID != configIDModel {
 		return nil, RPCError(codeInvalidParams, "unknown config option: "+p.ConfigID)
 	}
-	return SetSessionConfigOptionResult{ConfigOptions: a.configOptions(sess)}, nil
+	sess.setModel(p.Value)
+	return SetSessionConfigOptionResult{}, nil
 }
 
 func (a *Agent) handleZeroSetModel(_ context.Context, params json.RawMessage) (any, error) {
@@ -412,24 +356,8 @@ func (a *Agent) handleZeroSetModel(_ context.Context, params json.RawMessage) (a
 	if sess == nil {
 		return nil, RPCError(codeInvalidParams, "unknown session: "+p.SessionID)
 	}
-	model := strings.TrimSpace(p.Model)
-	if err := a.updateModel(sess, model, false); err != nil {
-		return nil, err
-	}
-	return ZeroSetModelResult{Model: model}, nil
-}
-
-func (a *Agent) updateModel(sess *acpSession, model string, restrictModels bool) error {
-	sess.modelMu.Lock()
-	defer sess.modelMu.Unlock()
-	if restrictModels && !sess.hasModel(model) {
-		return RPCError(codeInvalidParams, "unknown model: "+model)
-	}
-	if _, err := a.deps.Store.UpdateModel(sess.id, model); err != nil {
-		return RPCError(codeInternalError, "save model selection: "+err.Error())
-	}
-	sess.setModel(model)
-	return nil
+	sess.setModel(p.Model)
+	return ZeroSetModelResult{Model: p.Model}, nil
 }
 
 func (a *Agent) handleCancel(_ context.Context, params json.RawMessage) {
@@ -445,121 +373,42 @@ func (a *Agent) handleCancel(_ context.Context, params json.RawMessage) {
 // ---- advertising helpers ----
 
 func (a *Agent) modeState(s *acpSession) *SessionModeState {
-	// auto/ask/plan are offered over ACP; Unsafe is gated to the operator (see
-	// handleSetMode) so a client can't grant itself no-prompt host access. Plan
-	// only narrows what a client can do (read-only, no write/shell tools), so
-	// unlike Unsafe there is no elevation risk in letting a client select it.
+	// Only auto/ask are offered over ACP; Unsafe is gated to the operator (see
+	// handleSetMode) so a client can't grant itself no-prompt host access.
 	return &SessionModeState{
 		CurrentModeID: string(s.currentMode()),
 		AvailableModes: []SessionMode{
 			{ID: string(agent.PermissionModeAuto), Name: "Auto", Description: "Run safe tools automatically; ask before risky ones."},
 			{ID: string(agent.PermissionModeAsk), Name: "Ask", Description: "Ask before every tool that changes state."},
-			{ID: string(agent.PermissionModePlan), Name: "Plan", Description: "Read-only planning; write and shell tools are hidden."},
 		},
 	}
 }
 
-// resolveModelChoices advertises authenticated live models when discovery is
-// available. The configured model is always retained as the sole fallback.
-func (a *Agent) resolveModelChoices(ctx context.Context, cwd string) (string, []SessionConfigOptionValue, bool, error) {
-	resolved, err := a.deps.ResolveConfig(cwd, config.Overrides{})
-	if err != nil {
-		return "", nil, false, err
-	}
-	selected := strings.TrimSpace(resolved.Provider.Model)
-	options := make([]SessionConfigOptionValue, 0, 8)
-	seen := make(map[string]bool)
-	add := func(id, description string) {
-		id = strings.TrimSpace(id)
-		if id == "" || seen[id] {
-			return
-		}
-		seen[id] = true
-		options = append(options, SessionConfigOptionValue{Value: id, Name: id, Description: description})
-	}
-	add(selected, "")
-	descriptor, knownProvider := providercatalog.Get(resolved.Provider.CatalogID)
-	restrictModels := knownProvider && !descriptor.Custom
-	if a.deps.DiscoverModels != nil {
-		discovered, discoverErr := a.deps.DiscoverModels(ctx, resolved.Provider)
-		if ctx.Err() != nil {
-			return "", nil, false, ctx.Err()
-		}
-		if discoverErr == nil && len(discovered) > 0 {
-			for _, model := range discovered {
-				if providermodelcatalog.ModelIDAllowedForProvider(resolved.Provider.CatalogID, model.ID) {
-					add(model.ID, model.Description)
-				}
-			}
-		}
-	}
-	return selected, options, restrictModels, nil
-}
-
-func modelChoiceExists(models []SessionConfigOptionValue, model string) bool {
-	for _, option := range models {
-		if option.Value == model {
-			return true
-		}
-	}
-	return false
-}
-
-func (a *Agent) configOptions(s *acpSession) []SessionConfigOption {
-	model, models, mode := s.configState()
-	return []SessionConfigOption{{
-		ID:           configIDModel,
-		Name:         "Model",
-		Description:  "Model used for this session.",
-		Category:     configCategoryModel,
-		Type:         configOptionTypeSelect,
-		CurrentValue: model,
-		Options:      models,
-	}, {
-		ID:           configIDMode,
-		Name:         "Mode",
-		Description:  "Permission mode used for this session.",
-		Category:     configCategoryMode,
-		Type:         configOptionTypeSelect,
-		CurrentValue: string(mode),
-		Options: []SessionConfigOptionValue{
-			{Value: string(agent.PermissionModeAuto), Name: "Auto", Description: "Run safe tools automatically; ask before risky ones."},
-			{Value: string(agent.PermissionModeAsk), Name: "Ask", Description: "Ask before every tool that changes state."},
-			{Value: string(agent.PermissionModePlan), Name: "Plan", Description: "Read-only planning; write and shell tools are hidden."},
-		},
-	}}
-}
-
 // ---- persistence + continuity ----
 
-func (a *Agent) persistTurn(sess *acpSession, user, assistant string) error {
-	defer sess.appendHistory(turnRecord{user: user, assistant: assistant})
+func (a *Agent) persistTurn(sess *acpSession, user, assistant string) {
+	if a.deps.Store != nil {
+		_, _ = a.deps.Store.AppendEvent(sess.id, sessions.AppendEventInput{
+			Type:    sessions.EventMessage,
+			Payload: map[string]any{"role": "user", "content": user},
+		})
+		if assistant != "" {
+			_, _ = a.deps.Store.AppendEvent(sess.id, sessions.AppendEventInput{
+				Type:    sessions.EventMessage,
+				Payload: map[string]any{"role": "assistant", "content": assistant},
+			})
+		}
+	}
+	sess.appendHistory(turnRecord{user: user, assistant: assistant})
+}
+
+func (a *Agent) loadHistory(sessionID string) []turnRecord {
 	if a.deps.Store == nil {
 		return nil
 	}
-	events := []sessions.AppendEventInput{
-		{
-			Type:    sessions.EventMessage,
-			Payload: map[string]any{"role": "user", "content": user},
-		},
-	}
-	if assistant != "" {
-		events = append(events, sessions.AppendEventInput{
-			Type:    sessions.EventMessage,
-			Payload: map[string]any{"role": "assistant", "content": assistant},
-		})
-	}
-	_, err := a.deps.Store.AppendEvents(sess.id, events)
-	return err
-}
-
-func (a *Agent) loadHistory(sessionID string) ([]turnRecord, error) {
-	if a.deps.Store == nil {
-		return nil, nil
-	}
 	events, err := a.deps.Store.ReadEvents(sessionID)
 	if err != nil {
-		return nil, err
+		return nil
 	}
 	var records []turnRecord
 	var pendingUser string
@@ -595,21 +444,7 @@ func (a *Agent) loadHistory(sessionID string) ([]turnRecord, error) {
 	if havePending {
 		records = append(records, turnRecord{user: pendingUser})
 	}
-	return records, nil
-}
-
-func (a *Agent) warnPersistence(note *notifier, action string, message string, err error) {
-	if err == nil {
-		return
-	}
-	sessionID := ""
-	if note != nil {
-		sessionID = note.sessionID
-	}
-	log.Printf("zero acp: failed to %s for session %s: %v", action, sessionID, err)
-	if note != nil {
-		note.text("\n\n[zero warning] " + message + "\n")
-	}
+	return records
 }
 
 // buildPrompt prepends prior conversation as context, since agent.Run drives a
@@ -657,13 +492,13 @@ func promptImages(blocks []ContentBlock) []zeroruntime.ImageBlock {
 // session is returned unchanged rather than orphaning its turn or resetting its
 // mode/model. history is set BEFORE publishing so no concurrent prompt can read a
 // half-initialized session.
-func (a *Agent) registerSession(id, cwd string, history []turnRecord, model string, models []SessionConfigOptionValue, restrictModels bool) *acpSession {
+func (a *Agent) registerSession(id, cwd string, history []turnRecord) *acpSession {
 	a.mu.Lock()
 	defer a.mu.Unlock()
 	if existing := a.sessions[id]; existing != nil {
 		return existing
 	}
-	sess := &acpSession{id: id, cwd: cwd, mode: agent.PermissionModeAuto, model: model, models: models, restrictModels: restrictModels, history: history}
+	sess := &acpSession{id: id, cwd: cwd, mode: agent.PermissionModeAuto, history: history}
 	a.sessions[id] = sess
 	return sess
 }
@@ -702,18 +537,7 @@ func (s *acpSession) currentMode() agent.PermissionMode {
 }
 
 func (s *acpSession) setModel(model string) {
-	model = strings.TrimSpace(model)
 	s.mu.Lock()
-	found := false
-	for _, option := range s.models {
-		if option.Value == model {
-			found = true
-			break
-		}
-	}
-	if !found && model != "" {
-		s.models = append(s.models, SessionConfigOptionValue{Value: model, Name: model})
-	}
 	s.model = model
 	s.mu.Unlock()
 }
@@ -724,24 +548,6 @@ func (s *acpSession) currentModel() string {
 	return s.model
 }
 
-func (s *acpSession) hasModel(model string) bool {
-	model = strings.TrimSpace(model)
-	s.mu.Lock()
-	defer s.mu.Unlock()
-	for _, option := range s.models {
-		if option.Value == model {
-			return true
-		}
-	}
-	return false
-}
-
-func (s *acpSession) configState() (string, []SessionConfigOptionValue, agent.PermissionMode) {
-	s.mu.Lock()
-	defer s.mu.Unlock()
-	return s.model, append([]SessionConfigOptionValue(nil), s.models...), s.mode
-}
-
 func (s *acpSession) appendHistory(rec turnRecord) {
 	s.mu.Lock()
 	s.history = append(s.history, rec)

internal/acp/agent_test.go

@@ -3,18 +3,14 @@ package acp
 import (
 	"context"
 	"encoding/json"
-	"errors"
 	"fmt"
 	"io"
-	"os"
-	"path/filepath"
 	"strings"
 	"testing"
 	"time"
 
 	"github.com/Gitlawb/zero/internal/agent"
 	"github.com/Gitlawb/zero/internal/config"
-	"github.com/Gitlawb/zero/internal/providermodeldiscovery"
 	"github.com/Gitlawb/zero/internal/sandbox"
 	"github.com/Gitlawb/zero/internal/sessions"
 	"github.com/Gitlawb/zero/internal/tools"
@@ -139,12 +135,6 @@ func TestACPEndToEndPrompt(t *testing.T) {
 	if newRes.Modes == nil || newRes.Modes.CurrentModeID != string(agent.PermissionModeAuto) {
 		t.Fatalf("expected auto mode, got %+v", newRes.Modes)
 	}
-	if len(newRes.ConfigOptions) != 2 || newRes.ConfigOptions[0].ID != configIDModel || newRes.ConfigOptions[0].CurrentValue != "fake-model" {
-		t.Fatalf("model config option = %+v, want fake-model fallback", newRes.ConfigOptions)
-	}
-	if newRes.ConfigOptions[1].ID != configIDMode || newRes.ConfigOptions[1].CurrentValue != string(agent.PermissionModeAuto) {
-		t.Fatalf("mode config option = %+v", newRes.ConfigOptions[1])
-	}
 
 	// session/prompt
 	var promptRes PromptResult
@@ -164,201 +154,6 @@ func TestACPEndToEndPrompt(t *testing.T) {
 	}
 }
 
-func TestACPModelConfigOptionsCatalogSelectionAndLoad(t *testing.T) {
-	deps := testDeps(t)
-	deps.ResolveConfig = func(_ string, o config.Overrides) (config.ResolvedConfig, error) {
-		model := "gpt-5.5"
-		if o.Provider.Model != "" {
-			model = o.Provider.Model
-		}
-		return config.ResolvedConfig{Provider: config.ProviderProfile{
-			Name: "ChatGPT", CatalogID: "chatgpt", Model: model,
-		}}, nil
-	}
-	deps.DiscoverModels = func(_ context.Context, _ config.ProviderProfile) ([]providermodeldiscovery.Model, error) {
-		return []providermodeldiscovery.Model{
-			{ID: " gpt-5.4-mini ", Description: "Fast"},
-			{ID: "gpt-5.5", Description: "duplicate configured model"},
-		}, nil
-	}
-	h := newHarness(t, deps)
-	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
-	defer cancel()
-
-	var created NewSessionResult
-	if err := h.client.Call(ctx, MethodSessionNew, NewSessionParams{Cwd: t.TempDir()}, &created); err != nil {
-		t.Fatalf("session/new: %v", err)
-	}
-	option := created.ConfigOptions[0]
-	if option.CurrentValue != "gpt-5.5" || len(option.Options) < 2 {
-		t.Fatalf("new model option = %+v", option)
-	}
-	for _, choice := range option.Options {
-		if choice.Name != choice.Value {
-			t.Fatalf("model choice name = %q, want model id %q", choice.Name, choice.Value)
-		}
-	}
-	var selected SetSessionConfigOptionResult
-	if err := h.client.Call(ctx, MethodSessionSetConfigOption, SetSessionConfigOptionParams{
-		SessionID: created.SessionID, ConfigID: configIDModel, Value: "  gpt-5.4-mini  ",
-	}, &selected); err != nil {
-		t.Fatalf("set_config_option: %v", err)
-	}
-	if got := selected.ConfigOptions[0].CurrentValue; got != "gpt-5.4-mini" {
-		t.Fatalf("selected model = %q", got)
-	}
-	if err := h.client.Call(ctx, MethodSessionSetConfigOption, SetSessionConfigOptionParams{
-		SessionID: created.SessionID, ConfigID: configIDModel, Value: "not-advertised",
-	}, &SetSessionConfigOptionResult{}); err == nil {
-		t.Fatal("unknown standard model selection was accepted")
-	}
-	h.stop()
-	h = newHarness(t, deps)
-	defer h.stop()
-	var loaded LoadSessionResult
-	if err := h.client.Call(ctx, MethodSessionLoad, LoadSessionParams{SessionID: created.SessionID}, &loaded); err != nil {
-		t.Fatalf("session/load: %v", err)
-	}
-	if len(loaded.ConfigOptions) != 2 || loaded.ConfigOptions[0].CurrentValue != "gpt-5.4-mini" {
-		t.Fatalf("load model option = %+v", loaded.ConfigOptions)
-	}
-	if loaded.ConfigOptions[1].CurrentValue != string(agent.PermissionModeAuto) {
-		t.Fatalf("load mode option = %+v", loaded.ConfigOptions[1])
-	}
-}
-
-func TestACPModelDiscoveryFailureUsesConfiguredFallbackOnly(t *testing.T) {
-	deps := testDeps(t)
-	deps.ResolveConfig = func(_ string, _ config.Overrides) (config.ResolvedConfig, error) {
-		return config.ResolvedConfig{Provider: config.ProviderProfile{
-			Name: "ChatGPT", CatalogID: "chatgpt", Model: " configured-model ",
-		}}, nil
-	}
-	deps.DiscoverModels = func(context.Context, config.ProviderProfile) ([]providermodeldiscovery.Model, error) {
-		return nil, fmt.Errorf("discovery unavailable")
-	}
-	h := newHarness(t, deps)
-	defer h.stop()
-	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
-	defer cancel()
-
-	var created NewSessionResult
-	if err := h.client.Call(ctx, MethodSessionNew, NewSessionParams{Cwd: t.TempDir()}, &created); err != nil {
-		t.Fatalf("session/new: %v", err)
-	}
-	option := created.ConfigOptions[0]
-	if option.CurrentValue != "configured-model" || len(option.Options) != 1 || option.Options[0].Value != "configured-model" {
-		t.Fatalf("fallback model option = %+v", option)
-	}
-}
-
-func TestACPCustomProviderAllowsUnadvertisedModel(t *testing.T) {
-	deps := testDeps(t)
-	deps.ResolveConfig = func(_ string, o config.Overrides) (config.ResolvedConfig, error) {
-		model := "configured-model"
-		if o.Provider.Model != "" {
-			model = o.Provider.Model
-		}
-		return config.ResolvedConfig{Provider: config.ProviderProfile{
-			Name: "Custom", CatalogID: "custom-openai-compatible", Model: model,
-		}}, nil
-	}
-	h := newHarness(t, deps)
-	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
-	defer cancel()
-
-	var created NewSessionResult
-	if err := h.client.Call(ctx, MethodSessionNew, NewSessionParams{Cwd: t.TempDir()}, &created); err != nil {
-		t.Fatalf("session/new: %v", err)
-	}
-	var selected SetSessionConfigOptionResult
-	if err := h.client.Call(ctx, MethodSessionSetConfigOption, SetSessionConfigOptionParams{
-		SessionID: created.SessionID, ConfigID: configIDModel, Value: " vendor/model ",
-	}, &selected); err != nil {
-		t.Fatalf("set custom model: %v", err)
-	}
-	if got := selected.ConfigOptions[0].CurrentValue; got != "vendor/model" {
-		t.Fatalf("custom model = %q", got)
-	}
-	h.stop()
-	h = newHarness(t, deps)
-	defer h.stop()
-	var loaded LoadSessionResult
-	if err := h.client.Call(ctx, MethodSessionLoad, LoadSessionParams{SessionID: created.SessionID}, &loaded); err != nil {
-		t.Fatalf("session/load: %v", err)
-	}
-	option := loaded.ConfigOptions[0]
-	if option.CurrentValue != "vendor/model" || !modelChoiceExists(option.Options, "vendor/model") {
-		t.Fatalf("loaded custom model option = %+v", option)
-	}
-}
-
-func TestACPModelDiscoveryFiltersProviderIncompatibleModels(t *testing.T) {
-	a := &Agent{deps: Deps{
-		ResolveConfig: func(string, config.Overrides) (config.ResolvedConfig, error) {
-			return config.ResolvedConfig{Provider: config.ProviderProfile{
-				CatalogID: "opencode-go-anthropic-compatible", Model: "minimax-m3",
-			}}, nil
-		},
-		DiscoverModels: func(context.Context, config.ProviderProfile) ([]providermodeldiscovery.Model, error) {
-			return []providermodeldiscovery.Model{{ID: "qwen3.7-plus"}, {ID: "claude-sonnet-4.5"}}, nil
-		},
-	}}
-	_, options, restricted, err := a.resolveModelChoices(context.Background(), t.TempDir())
-	if err != nil {
-		t.Fatal(err)
-	}
-	if !restricted || !modelChoiceExists(options, "qwen3.7-plus") || modelChoiceExists(options, "claude-sonnet-4.5") {
-		t.Fatalf("filtered model options = %+v, restricted=%v", options, restricted)
-	}
-}
-
-func TestACPModelDiscoveryHonorsCancellation(t *testing.T) {
-	ctx, cancel := context.WithCancel(context.Background())
-	cancel()
-	a := &Agent{deps: Deps{
-		ResolveConfig: func(string, config.Overrides) (config.ResolvedConfig, error) {
-			return config.ResolvedConfig{Provider: config.ProviderProfile{Model: "configured-model"}}, nil
-		},
-		DiscoverModels: func(context.Context, config.ProviderProfile) ([]providermodeldiscovery.Model, error) {
-			return nil, context.Canceled
-		},
-	}}
-	if _, _, _, err := a.resolveModelChoices(ctx, t.TempDir()); !errors.Is(err, context.Canceled) {
-		t.Fatalf("resolve error = %v, want context canceled", err)
-	}
-}
-
-func TestACPConfigOptionWireSchema(t *testing.T) {
-	b, err := json.Marshal(SessionConfigOption{
-		ID: "model", Name: "Model", Description: "desc", Category: "model",
-		Type: configOptionTypeSelect, CurrentValue: "m1",
-		Options: []SessionConfigOptionValue{{Value: "m1", Name: "m1", Description: "choice"}},
-	})
-	if err != nil {
-		t.Fatal(err)
-	}
-	var wire map[string]any
-	if err := json.Unmarshal(b, &wire); err != nil {
-		t.Fatal(err)
-	}
-	for _, key := range []string{"type", "category", "currentValue", "options"} {
-		if _, ok := wire[key]; !ok {
-			t.Errorf("wire field %q absent: %s", key, b)
-		}
-	}
-	if _, ok := wire["value"]; ok {
-		t.Errorf("obsolete option value field present: %s", b)
-	}
-	if _, ok := wire["values"]; ok {
-		t.Errorf("obsolete values field present: %s", b)
-	}
-	choice := wire["options"].([]any)[0].(map[string]any)
-	if choice["value"] != "m1" {
-		t.Errorf("options[].value = %#v", choice["value"])
-	}
-}
-
 func TestACPUnknownSessionPromptErrors(t *testing.T) {
 	h := newHarness(t, testDeps(t))
 	defer h.stop()
@@ -384,41 +179,6 @@ func TestACPSetModeUpdatesSession(t *testing.T) {
 	if err := h.client.Call(ctx, MethodSessionSetMode, SetSessionModeParams{SessionID: newRes.SessionID, ModeID: string(agent.PermissionModeAsk)}, &SetSessionModeResult{}); err != nil {
 		t.Fatalf("set_mode ask: %v", err)
 	}
-	var configured SetSessionConfigOptionResult
-	if err := h.client.Call(ctx, MethodSessionSetConfigOption, SetSessionConfigOptionParams{
-		SessionID: newRes.SessionID, ConfigID: configIDMode, Value: string(agent.PermissionModeAuto),
-	}, &configured); err != nil {
-		t.Fatalf("set_config_option mode: %v", err)
-	}
-	if got := configured.ConfigOptions[1].CurrentValue; got != string(agent.PermissionModeAuto) {
-		t.Fatalf("configured mode = %q", got)
-	}
-	// Plan is accepted: it only narrows capability (read-only), so unlike Unsafe
-	// there is no elevation risk in letting a client select it.
-	if err := h.client.Call(ctx, MethodSessionSetMode, SetSessionModeParams{SessionID: newRes.SessionID, ModeID: string(agent.PermissionModePlan)}, &SetSessionModeResult{}); err != nil {
-		t.Fatalf("set_mode plan: %v", err)
-	}
-	// Configuration path is a separate contract from MethodSessionSetMode: the
-	// mode option is advertised via configOptions and applied by handleSetConfigOption.
-	var planConfigured SetSessionConfigOptionResult
-	if err := h.client.Call(ctx, MethodSessionSetConfigOption, SetSessionConfigOptionParams{
-		SessionID: newRes.SessionID, ConfigID: configIDMode, Value: string(agent.PermissionModePlan),
-	}, &planConfigured); err != nil {
-		t.Fatalf("set_config_option plan: %v", err)
-	}
-	if got := planConfigured.ConfigOptions[1].CurrentValue; got != string(agent.PermissionModePlan) {
-		t.Fatalf("configured mode after plan = %q, want plan", got)
-	}
-	hasPlanOption := false
-	for _, opt := range planConfigured.ConfigOptions[1].Options {
-		if opt.Value == string(agent.PermissionModePlan) {
-			hasPlanOption = true
-			break
-		}
-	}
-	if !hasPlanOption {
-		t.Fatalf("config mode options missing plan: %#v", planConfigured.ConfigOptions[1].Options)
-	}
 	// Unsafe must be rejected over ACP — a client can't self-grant no-prompt host access.
 	if err := h.client.Call(ctx, MethodSessionSetMode, SetSessionModeParams{SessionID: newRes.SessionID, ModeID: string(agent.PermissionModeUnsafe)}, &SetSessionModeResult{}); err == nil {
 		t.Fatal("expected Unsafe mode to be rejected over ACP")
@@ -429,37 +189,6 @@ func TestACPSetModeUpdatesSession(t *testing.T) {
 	}
 }
 
-// TestACPPlanModeWiresPermissionModeIntoAgentOptions confirms selecting "plan"
-// over ACP actually reaches agent.Options.PermissionMode for the next turn —
-// the same gap this test's TUI counterpart covers for /plan on.
-func TestACPPlanModeWiresPermissionModeIntoAgentOptions(t *testing.T) {
-	deps := testDeps(t)
-	var captured agent.Options
-	deps.RunAgent = func(_ context.Context, _ string, _ zeroruntime.Provider, opts agent.Options) (agent.Result, error) {
-		captured = opts
-		return agent.Result{FinalAnswer: "ok"}, nil
-	}
-
-	h := newHarness(t, deps)
-	defer h.stop()
-	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
-	defer cancel()
-
-	var newRes NewSessionResult
-	if err := h.client.Call(ctx, MethodSessionNew, NewSessionParams{Cwd: t.TempDir(), McpServers: []McpServer{}}, &newRes); err != nil {
-		t.Fatalf("session/new: %v", err)
-	}
-	if err := h.client.Call(ctx, MethodSessionSetMode, SetSessionModeParams{SessionID: newRes.SessionID, ModeID: string(agent.PermissionModePlan)}, &SetSessionModeResult{}); err != nil {
-		t.Fatalf("set_mode plan: %v", err)
-	}
-	if err := h.client.Call(ctx, MethodSessionPrompt, PromptParams{SessionID: newRes.SessionID, Prompt: []ContentBlock{TextBlock("plan it out")}}, &PromptResult{}); err != nil {
-		t.Fatalf("session/prompt: %v", err)
-	}
-	if captured.PermissionMode != agent.PermissionModePlan {
-		t.Fatalf("agent.Options.PermissionMode = %q, want plan", captured.PermissionMode)
-	}
-}
-
 // TestACPRunTurnWiresSandboxAndScopedRegistry proves the sandbox engine and the
 // scoped registry from BuildWorkspace actually reach agent.Options — i.e. ACP
 // shell tools run confined, not unconfined on the host.
@@ -513,77 +242,8 @@ func TestACPRejectsInvalidCwd(t *testing.T) {
 	}
 }
 
-func TestACPPromptWarnsWhenTurnPersistenceFails(t *testing.T) {
-	deps := testDeps(t)
-	h := newHarness(t, deps)
-	defer h.stop()
-	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
-	defer cancel()
-
-	var newRes NewSessionResult
-	if err := h.client.Call(ctx, MethodSessionNew, NewSessionParams{Cwd: t.TempDir(), McpServers: []McpServer{}}, &newRes); err != nil {
-		t.Fatalf("session/new: %v", err)
-	}
-	metadataPath := filepath.Join(deps.Store.RootDir, newRes.SessionID, sessions.MetadataFile)
-	if err := os.Remove(metadataPath); err != nil {
-		t.Fatalf("remove metadata: %v", err)
-	}
-
-	var promptRes PromptResult
-	if err := h.client.Call(ctx, MethodSessionPrompt, PromptParams{
-		SessionID: newRes.SessionID,
-		Prompt:    []ContentBlock{TextBlock("hi")},
-	}, &promptRes); err != nil {
-		t.Fatalf("session/prompt: %v", err)
-	}
-	if promptRes.StopReason != StopEndTurn {
-		t.Fatalf("stopReason = %q, want %q", promptRes.StopReason, StopEndTurn)
-	}
-	got := drainTextUntil(t, h.updates, func(text string) bool {
-		return strings.Contains(text, "Hello from ZERO") &&
-			strings.Contains(text, "Could not save session history")
-	})
-	if !strings.Contains(got, "Could not save session history") {
-		t.Fatalf("streamed text = %q, want persistence warning", got)
-	}
-}
-
-func TestACPLoadWarnsWhenHistoryReadFails(t *testing.T) {
-	deps := testDeps(t)
-	cwd := t.TempDir()
-	meta, err := deps.Store.Create(sessions.CreateInput{Title: "ACP session", Cwd: cwd})
-	if err != nil {
-		t.Fatalf("create session: %v", err)
-	}
-	eventsPath := filepath.Join(deps.Store.RootDir, meta.SessionID, sessions.EventsFile)
-	if err := os.WriteFile(eventsPath, []byte("{bad json}\n"), 0o600); err != nil {
-		t.Fatalf("write corrupt events: %v", err)
-	}
-	h := newHarness(t, deps)
-	defer h.stop()
-	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
-	defer cancel()
-
-	if err := h.client.Call(ctx, MethodSessionLoad, LoadSessionParams{SessionID: meta.SessionID, Cwd: cwd, McpServers: []McpServer{}}, &LoadSessionResult{}); err != nil {
-		t.Fatalf("session/load: %v", err)
-	}
-	got := drainTextUntil(t, h.updates, func(text string) bool {
-		return strings.Contains(text, "Could not load session history")
-	})
-	if !strings.Contains(got, "Could not load session history") {
-		t.Fatalf("streamed text = %q, want load warning", got)
-	}
-}
-
 // drainText collects streamed chunks for a short window and concatenates them.
 func drainText(t *testing.T, ch <-chan string) string {
-	t.Helper()
-	return drainTextUntil(t, ch, func(text string) bool {
-		return strings.Contains(text, "Hello from ZERO")
-	})
-}
-
-func drainTextUntil(t *testing.T, ch <-chan string, done func(string) bool) string {
 	t.Helper()
 	var b strings.Builder
 	deadline := time.After(2 * time.Second)
@@ -591,7 +251,7 @@ func drainTextUntil(t *testing.T, ch <-chan string, done func(string) bool) stri
 		select {
 		case s := <-ch:
 			b.WriteString(s)
-			if done(b.String()) {
+			if strings.Contains(b.String(), "Hello from ZERO") {
 				return b.String()
 			}
 		case <-deadline:

internal/acp/export_test.go

@@ -1,11 +0,0 @@
-// Test seams: helpers only test code uses, kept out of the production binary.
-package acp
-
-import "context"
-
-// Serve runs the connection read loop until the stream closes or ctx is done.
-func (a *Agent) Serve(ctx context.Context) error { return a.conn.Serve(ctx) }
-
-func ImageBlock(base64Data, mimeType string) ContentBlock {
-	return ContentBlock{Type: "image", Data: base64Data, MimeType: mimeType}
-}

internal/acp/jsonrpc.go

@@ -17,7 +17,6 @@ import (
 	"fmt"
 	"io"
 	"sync"
-	"sync/atomic"
 )
 
 // JSON-RPC 2.0 standard error codes.
@@ -81,9 +80,8 @@ type NotifyFunc func(ctx context.Context, params json.RawMessage)
 // requests/notifications — needed because ACP is bidirectional (the agent calls
 // the client for session/request_permission, fs/*, terminal/*).
 type Conn struct {
-	rawReader    io.Reader // wrapped lazily in Serve, once ctx is known — see interruptibleReader
-	readerCloser io.Closer
-	w            io.Writer
+	reader *bufio.Reader
+	w      io.Writer
 
 	writeMu sync.Mutex // serializes all writes to w
 
@@ -98,15 +96,10 @@ type Conn struct {
 	wg sync.WaitGroup // tracks in-flight inbound handlers
 }
 
-// NewConn builds a peer reading ndjson from r and writing ndjson to w. This
-// Conn does not own r: cancelling Serve's context cannot forcibly close it, so
-// a Read on r that's genuinely idle (never returns on its own) leaves Serve
-// blocked until r itself produces something — e.g. a test that closes r's own
-// write side to unblock it. Use NewOwnedConn when this Conn should own r's
-// lifetime.
+// NewConn builds a peer reading ndjson from r and writing ndjson to w.
 func NewConn(r io.Reader, w io.Writer) *Conn {
 	return &Conn{
-		rawReader: r,
+		reader:    bufio.NewReader(r),
 		w:         w,
 		handlers:  make(map[string]HandlerFunc),
 		notifiers: make(map[string]NotifyFunc),
@@ -114,18 +107,6 @@ func NewConn(r io.Reader, w io.Writer) *Conn {
 	}
 }
 
-// NewOwnedConn is like NewConn but declares that this Conn exclusively owns r:
-// if r implements io.Closer, Serve closes it on context cancellation to
-// interrupt an otherwise-idle blocking Read. Only pass a reader whose lifetime
-// this Conn should control — e.g. the process's own stdin — never a reader a
-// caller retains or shares elsewhere, since cancellation tears it down
-// permanently for every consumer, not just this Conn.
-func NewOwnedConn(r io.Reader, w io.Writer) *Conn {
-	c := NewConn(r, w)
-	c.readerCloser, _ = r.(io.Closer)
-	return c
-}
-
 // Handle registers a request handler for method.
 func (c *Conn) Handle(method string, fn HandlerFunc) { c.handlers[method] = fn }
 
@@ -147,36 +128,17 @@ func (c *Conn) Serve(ctx context.Context) error {
 		c.wg.Wait()
 	}()
 
-	interruptible := newInterruptibleReader(ctx, c.rawReader, c.readerCloser)
-	reader := bufio.NewReader(interruptible)
-
 	for {
 		// One ndjson value per line. Framing per line (rather than streaming the
 		// decoder) keeps a single malformed line from making the whole connection
 		// unrecoverable — we report -32700 and continue.
-		//
-		// generation brackets this SPECIFIC ReadBytes call: bufio may invoke
-		// interruptible.Read zero or more times underneath it (zero when the
-		// answer is already sitting in bufio's own buffer — including a
-		// previously-buffered, not-yet-surfaced error: bufio can return a
-		// complete line with a nil error while quietly caching an error it
-		// received ALONGSIDE that line's bytes, which then surfaces on the VERY
-		// NEXT call with no further contact with interruptible.Read at all). The
-		// generation only advances inside interruptible.Read's own cancellation
-		// branch, so "unchanged across this call" is proof this call's outcome
-		// didn't pass through that branch — whether because it was answered from
-		// the buffer or because it raced ctx.Done() and won. Only that proof
-		// makes it safe to call the outcome genuine.
-		before := interruptible.generation()
-		line, err := reader.ReadBytes('\n')
-		interrupted := interruptible.generation() != before
-
+		line, err := c.reader.ReadBytes('\n')
 		if len(bytes.TrimSpace(line)) > 0 {
 			c.handleLine(ctx, line)
 		}
 		if err != nil {
 			c.failAllPending(err)
-			if errors.Is(err, io.EOF) || interrupted {
+			if errors.Is(err, io.EOF) || ctx.Err() != nil {
 				return nil
 			}
 			return err
@@ -184,100 +146,6 @@ func (c *Conn) Serve(ctx context.Context) error {
 	}
 }
 
-// interruptibleReader wraps a reader so a context cancellation can unblock a
-// currently in-flight Read without ever discarding — or misattributing — the
-// real outcome of a call that wasn't actually blocked.
-//
-// Each Read runs the underlying call on its own goroutine and races it against
-// ctx.Done(). If the underlying call finishes first, its result is returned
-// exactly as received; ctx being cancelled a moment later has no bearing on a
-// call that already had its answer. If ctx.Done() fires first, closer (when
-// non-nil) is closed to nudge a genuinely blocked read, the generation counter
-// is bumped, and the call STILL waits for and returns the underlying call's own
-// result — never a synthesized one — since closing a reader's outcome is
-// whatever the reader itself reports for it, not ours to invent.
-//
-// The counter is what lets Serve tell a call that was actually raced against
-// cancellation apart from one merely answered while cancellation happened to be
-// pending — the distinction the read loop's own doc comment explains.
-type interruptibleReader struct {
-	inner  io.Reader
-	closer io.Closer
-	ctx    context.Context
-
-	closeOnce sync.Once
-	gen       atomic.Int64
-}
-
-func newInterruptibleReader(ctx context.Context, inner io.Reader, closer io.Closer) *interruptibleReader {
-	return &interruptibleReader{inner: inner, closer: closer, ctx: ctx}
-}
-
-func (r *interruptibleReader) generation() int64 { return r.gen.Load() }
-
-type interruptibleReadResult struct {
-	n   int
-	err error
-}
-
-// interruptibleReadDecision arbitrates whether an underlying Read completed or
-// cancellation claimed it first. Completion is recorded before the result is
-// published so cancellation cannot misclassify a finished read merely because
-// its goroutine has not sent the result yet.
-type interruptibleReadDecision struct {
-	once sync.Once
-}
-
-func (d *interruptibleReadDecision) complete() { d.once.Do(func() {}) }
-
-func (d *interruptibleReadDecision) interrupt() (claimed bool) {
-	d.once.Do(func() { claimed = true })
-	return claimed
-}
-
-// readInner records completion in the Read call's deferred epilogue. It also
-// converts a reader panic into a terminal error so the helper goroutine always
-// publishes a result for Read to join.
-func (r *interruptibleReader) readInner(p []byte, decision *interruptibleReadDecision) (res interruptibleReadResult) {
-	defer func() {
-		if recovered := recover(); recovered != nil {
-			res = interruptibleReadResult{err: fmt.Errorf("acp: reader panicked: %v", recovered)}
-		}
-		decision.complete()
-	}()
-	res.n, res.err = r.inner.Read(p)
-	return res
-}
-
-func (r *interruptibleReader) Read(p []byte) (int, error) {
-	resultCh := make(chan interruptibleReadResult, 1)
-	var decision interruptibleReadDecision
-	go func() {
-		resultCh <- r.readInner(p, &decision)
-	}()
-	select {
-	case res := <-resultCh:
-		return res.n, res.err
-	case <-r.ctx.Done():
-		// resultCh may not be readable yet even when inner.Read has returned: its
-		// goroutine can be descheduled between recording completion and publishing
-		// the result. Only interrupt a call that cancellation actually claims.
-		if decision.interrupt() {
-			r.gen.Add(1)
-			if r.closer != nil {
-				r.closeOnce.Do(func() { _ = r.closer.Close() })
-			}
-		}
-		// Wait for the real result rather than returning synthetically: p is
-		// still being written by the goroutine above until it does, so returning
-		// early would race that write, and the actual error (whatever closing
-		// the reader produced, or whatever the reader was already about to
-		// report) is what the caller should see.
-		res := <-resultCh
-		return res.n, res.err
-	}
-}
-
 // handleLine decodes and dispatches one ndjson frame. A parse failure replies
 // with -32700 (id null) and keeps the connection alive.
 func (c *Conn) handleLine(ctx context.Context, line []byte) {

internal/acp/jsonrpc_test.go

@@ -3,30 +3,11 @@ package acp
 import (
 	"context"
 	"encoding/json"
-	"errors"
 	"io"
-	"sync"
-	"sync/atomic"
 	"testing"
 	"time"
 )
 
-type testReadCloser struct {
-	read  func([]byte) (int, error)
-	close func() error
-}
-
-func (r testReadCloser) Read(p []byte) (int, error) { return r.read(p) }
-func (r testReadCloser) Close() error               { return r.close() }
-
-type testReader func([]byte) (int, error)
-
-func (r testReader) Read(p []byte) (int, error) { return r(p) }
-
-type testWriter func([]byte) (int, error)
-
-func (w testWriter) Write(p []byte) (int, error) { return w(p) }
-
 // connPair wires two Conns together over in-memory pipes and serves both.
 func connPair(t *testing.T) (a, b *Conn, stop func()) {
 	t.Helper()
@@ -44,275 +25,6 @@ func connPair(t *testing.T) (a, b *Conn, stop func()) {
 	}
 }
 
-func TestConnServeCancellationInterruptsIdleRead(t *testing.T) {
-	pipeReader, writer := io.Pipe()
-	t.Cleanup(func() { _ = writer.Close() })
-	// Buffered + non-blocking send, not close: interruptibleReader retries each
-	// underlying Read on its own goroutine, so if bufio ever calls this more than
-	// once, a close() here would panic on the second call.
-	readStarted := make(chan struct{}, 1)
-	closeCalls := make(chan struct{}, 2)
-	reader := testReadCloser{
-		read: func(p []byte) (int, error) {
-			select {
-			case readStarted <- struct{}{}:
-			default:
-			}
-			return pipeReader.Read(p)
-		},
-		close: func() error {
-			closeCalls <- struct{}{}
-			return pipeReader.Close()
-		},
-	}
-	conn := NewOwnedConn(reader, io.Discard)
-	ctx, cancel := context.WithCancel(context.Background())
-	done := make(chan error, 1)
-	go func() { done <- conn.Serve(ctx) }()
-
-	<-readStarted
-	cancel()
-	select {
-	case err := <-done:
-		if err != nil {
-			t.Fatalf("Serve returned %v after cancellation, want nil", err)
-		}
-	case <-time.After(time.Second):
-		t.Fatal("Serve did not return after cancelling an idle connection")
-	}
-	if got := len(closeCalls); got != 1 {
-		t.Fatalf("reader Close calls = %d, want 1", got)
-	}
-}
-
-func TestConnServePreservesTerminalReadErrorDuringCancellation(t *testing.T) {
-	for _, closable := range []bool{false, true} {
-		name := "non-closable"
-		if closable {
-			name = "closable"
-		}
-		t.Run(name, func(t *testing.T) {
-			wantErr := errors.New("read failed")
-			read := testReader(func(p []byte) (int, error) {
-				return copy(p, "not json"), wantErr
-			})
-			closeCalls := make(chan struct{}, 1)
-			var reader io.Reader = read
-			if closable {
-				reader = testReadCloser{
-					read: read,
-					close: func() error {
-						closeCalls <- struct{}{}
-						return nil
-					},
-				}
-			}
-			// Buffered + non-blocking send, not close: a later change adding a
-			// second write (e.g. an additional error frame) must not panic on a
-			// repeated close of an already-closed channel.
-			writeStarted := make(chan struct{}, 1)
-			releaseWrite := make(chan struct{})
-			writer := testWriter(func(p []byte) (int, error) {
-				select {
-				case writeStarted <- struct{}{}:
-				default:
-				}
-				<-releaseWrite
-				return len(p), nil
-			})
-			newConn := NewConn
-			if closable {
-				// Ownership matters here: this case exists to prove that even a
-				// Conn entitled to close its reader on cancellation still
-				// preserves a terminal error that was already decided.
-				newConn = NewOwnedConn
-			}
-			conn := newConn(reader, writer)
-			ctx, cancel := context.WithCancel(context.Background())
-			done := make(chan error, 1)
-			go func() { done <- conn.Serve(ctx) }()
-
-			<-writeStarted
-			cancel()
-			close(releaseWrite)
-			select {
-			case err := <-done:
-				if !errors.Is(err, wantErr) {
-					t.Fatalf("Serve returned %v, want terminal read error %v", err, wantErr)
-				}
-			case <-time.After(time.Second):
-				t.Fatal("Serve did not return after terminal read error")
-			}
-			if got := len(closeCalls); got != 0 {
-				t.Fatalf("reader Close calls = %d, want 0 after terminal read", got)
-			}
-		})
-	}
-}
-
-// TestConnServePreservesReadErrorBufferedAlongsideAPriorLine is the deterministic
-// regression test for jatmn's #782 finding: a real, non-EOF ReadBytes failure can
-// be DECIDED before cancellation ever happens, yet only SURFACE afterward, and
-// must still be reported rather than swallowed as a clean shutdown.
-//
-// bufio.Reader can obtain data AND a terminal error from ONE underlying Read call
-// (io.Reader explicitly permits returning (n>0, err) together). If a delimiter is
-// found within that data, ReadBytes returns the line with a nil error for THIS
-// call, but bufio caches the error internally and returns it — WITHOUT calling the
-// underlying reader again — on the very next ReadBytes call. This test forces that
-// exact sequence: the underlying reader hands back one complete, validly-framed
-// line together with wantErr in a SINGLE call (asserted below to be its ONLY
-// call), the loop dispatches that line through a synchronous path that blocks the
-// read loop, cancellation is asserted to have happened before the loop attempts
-// its next read, and only then is the block released. By the time the read loop
-// asks for more input, cancellation is already in effect and the underlying
-// reader is never touched again — so if Serve attributed the resulting error to
-// cancellation merely because ctx was already done, it would incorrectly return
-// nil. It must instead recognize that this read was answered entirely from
-// bufio's own cache, never raced against ctx.Done(), and report wantErr.
-func TestConnServePreservesReadErrorBufferedAlongsideAPriorLine(t *testing.T) {
-	wantErr := errors.New("read failed")
-	var readCalls atomic.Int32
-	read := testReader(func(p []byte) (int, error) {
-		readCalls.Add(1)
-		// An unsupported jsonrpc version with an id takes the SYNCHRONOUS
-		// writeError path in handleLine (not a dispatched goroutine), giving this
-		// test a reliable blocking point between processing this line and the
-		// loop's next ReadBytes call.
-		return copy(p, `{"jsonrpc":"1.0","id":1}`+"\n"), wantErr
-	})
-	writeStarted := make(chan struct{}, 1)
-	releaseWrite := make(chan struct{})
-	var releaseWriteOnce sync.Once
-	release := func() { releaseWriteOnce.Do(func() { close(releaseWrite) }) }
-	t.Cleanup(release)
-	writer := testWriter(func(p []byte) (int, error) {
-		select {
-		case writeStarted <- struct{}{}:
-		default:
-		}
-		<-releaseWrite
-		return len(p), nil
-	})
-	conn := NewConn(read, writer)
-	ctx, cancel := context.WithCancel(context.Background())
-	done := make(chan error, 1)
-	go func() { done <- conn.Serve(ctx) }()
-
-	// The synchronous write confirms the first (and, per the panic guard above,
-	// only) read has already returned and is being processed — cancelling now
-	// lands squarely in the gap between that read and the loop's next one, never
-	// racing an in-flight interruptibleReader.Read call.
-	select {
-	case <-writeStarted:
-	case <-time.After(time.Second):
-		t.Fatal("Serve did not reach the synchronous write")
-	}
-	cancel()
-	release()
-
-	select {
-	case err := <-done:
-		if !errors.Is(err, wantErr) {
-			t.Fatalf("Serve returned %v, want the buffered terminal read error %v", err, wantErr)
-		}
-	case <-time.After(time.Second):
-		t.Fatal("Serve did not return after the buffered terminal read error")
-	}
-	if got := readCalls.Load(); got != 1 {
-		t.Fatalf("underlying Read calls = %d, want exactly 1 (bufio must have served the second ReadBytes from its cache)", got)
-	}
-}
-
-// TestInterruptibleReaderPrefersDecidedResultOverCancellation is the regression
-// test for jatmn's second #782 finding: interruptibleReader.Read's select
-// between resultCh and ctx.Done() chooses pseudo-randomly when both are ready,
-// so a real, already-decided outcome (e.g. a terminal transport error) could be
-// misattributed to cancellation purely because ctx also happened to be done by
-// the time the select ran. Each iteration lets the underlying Read return (so
-// resultCh already holds the real outcome) before cancelling, maximizing the
-// odds of landing exactly in that ambiguous window; the generation counter must
-// never advance when the real, already-decided result is what gets returned.
-func TestInterruptibleReaderPrefersDecidedResultOverCancellation(t *testing.T) {
-	wantErr := errors.New("read failed")
-	for i := 0; i < 100; i++ {
-		resultReady := make(chan struct{})
-		read := testReader(func(p []byte) (int, error) {
-			defer close(resultReady)
-			return copy(p, "x"), wantErr
-		})
-		ctx, cancel := context.WithCancel(context.Background())
-		r := newInterruptibleReader(ctx, read, nil)
-
-		readDone := make(chan interruptibleReadResult, 1)
-		go func() {
-			n, err := r.Read(make([]byte, 8))
-			readDone <- interruptibleReadResult{n, err}
-		}()
-
-		<-resultReady
-		time.Sleep(time.Millisecond) // bias resultCh's send ahead of cancel
-		cancel()
-
-		res := <-readDone
-		if !errors.Is(res.err, wantErr) {
-			t.Fatalf("iteration %d: Read returned %v, want %v", i, res.err, wantErr)
-		}
-		if got := r.generation(); got != 0 {
-			t.Fatalf("iteration %d: generation = %d, want 0 (a decided result must never be attributed to cancellation)", i, got)
-		}
-		cancel()
-	}
-}
-
-// TestInterruptibleReadCompletionWinsBeforeResultPublication exercises the
-// actual inner-read wrapper at the handoff between the underlying Read and the
-// helper goroutine's result publication. Cancellation must not reclassify a
-// result returned by this wrapper as interrupted.
-func TestInterruptibleReadCompletionWinsBeforeResultPublication(t *testing.T) {
-	wantErr := errors.New("read failed")
-	read := testReader(func(p []byte) (int, error) {
-		return copy(p, "x"), wantErr
-	})
-	r := newInterruptibleReader(context.Background(), read, nil)
-	var decision interruptibleReadDecision
-	readComplete := make(chan struct{})
-	resultCh := make(chan interruptibleReadResult)
-	go func() {
-		res := r.readInner(make([]byte, 8), &decision)
-		close(readComplete)
-		resultCh <- res
-	}()
-
-	<-readComplete
-	if decision.interrupt() {
-		t.Fatal("cancellation claimed a read that completed before result publication")
-	}
-	res := <-resultCh
-	if res.n != 1 || !errors.Is(res.err, wantErr) {
-		t.Fatalf("inner read result = (%d, %v), want (1, %v)", res.n, res.err, wantErr)
-	}
-}
-
-func TestConnServeReportsReaderPanicAsTerminalError(t *testing.T) {
-	read := testReader(func([]byte) (int, error) {
-		panic("boom")
-	})
-	conn := NewConn(read, io.Discard)
-	done := make(chan error, 1)
-	go func() { done <- conn.Serve(context.Background()) }()
-
-	select {
-	case err := <-done:
-		const want = "acp: reader panicked: boom"
-		if err == nil || err.Error() != want {
-			t.Fatalf("Serve returned %v, want %q", err, want)
-		}
-	case <-time.After(time.Second):
-		t.Fatal("Serve did not return after reader panic")
-	}
-}
-
 func TestConnRequestResponse(t *testing.T) {
 	a, b, stop := connPair(t)
 	defer stop()

internal/acp/permission.go

@@ -25,7 +25,7 @@ func buildPermissionOptions(req agent.PermissionRequest) []PermissionOption {
 	}
 	options := make([]PermissionOption, 0, len(actions))
 	for _, action := range actions {
-		kind, name := optionKindFor(action, req.PrefixApprovalEscalates)
+		kind, name := optionKindFor(action)
 		if kind == "" {
 			continue // skip actions that have no clean ACP option (e.g. cancel)
 		}
@@ -41,24 +41,18 @@ func buildPermissionOptions(req agent.PermissionRequest) []PermissionOption {
 // optionKindFor maps a ZERO decision action to an ACP PermissionOptionKind and a
 // human label. Returns an empty kind for actions that ACP expresses through the
 // outcome rather than an option (cancel).
-//
-// escalates reports that approving a command prefix will also run the command
-// outside the sandbox. The two prefix options say so in their label, since the
-// label is all an ACP client shows and the option id carries no room for it. An
-// editor user is deciding on exactly the same terms as a terminal user, so the
-// disclosure cannot be TUI-only.
-func optionKindFor(action agent.PermissionDecisionAction, escalates bool) (kind, name string) {
+func optionKindFor(action agent.PermissionDecisionAction) (kind, name string) {
 	switch action {
 	case agent.PermissionDecisionAllow, agent.PermissionDecisionAllowStrict:
 		return PermAllowOnce, "Allow"
 	case agent.PermissionDecisionAllowForSession:
 		return PermAllowAlways, "Allow for this session"
 	case agent.PermissionDecisionAllowPrefix:
-		return PermAllowAlways, withSandboxEscalationNote("Allow this command for the session", escalates)
+		return PermAllowAlways, "Allow this command for the session"
 	case agent.PermissionDecisionAlwaysAllow:
 		return PermAllowAlways, "Always allow"
 	case agent.PermissionDecisionAlwaysAllowPrefix:
-		return PermAllowAlways, withSandboxEscalationNote("Always allow this command", escalates)
+		return PermAllowAlways, "Always allow this command"
 	case agent.PermissionDecisionDeny:
 		return PermRejectOnce, "Reject"
 	case agent.PermissionDecisionCancel:
@@ -129,13 +123,3 @@ func rawInputBytes(data []byte) json.RawMessage {
 	}
 	return json.RawMessage(data)
 }
-
-// withSandboxEscalationNote appends the escalation disclosure to a prefix
-// option's label. Kept parallel to the TUI's wording so the same consequence is
-// not described two different ways depending on which client the user runs.
-func withSandboxEscalationNote(name string, escalates bool) string {
-	if !escalates {
-		return name
-	}
-	return name + " (runs outside the sandbox)"
-}

internal/acp/permission_escalation_test.go

@@ -1,68 +0,0 @@
-package acp
-
-import (
-	"strings"
-	"testing"
-
-	"github.com/Gitlawb/zero/internal/agent"
-)
-
-func escalationRequest(escalates bool) agent.PermissionRequest {
-	return agent.PermissionRequest{
-		ToolName: "bash",
-		AvailableDecisions: []agent.PermissionDecisionAction{
-			agent.PermissionDecisionAllow,
-			agent.PermissionDecisionAllowPrefix,
-			agent.PermissionDecisionAlwaysAllowPrefix,
-			agent.PermissionDecisionDeny,
-		},
-		PrefixApprovalEscalates: escalates,
-	}
-}
-
-// An ACP client renders only Name, so if the escalation is not in the label the
-// editor user approves it blind. They face exactly the outcome a terminal user
-// faces, so the disclosure has to reach both clients or it is not a fix.
-func TestACPPrefixOptionsDiscloseLeavingTheSandbox(t *testing.T) {
-	found := 0
-	for _, option := range buildPermissionOptions(escalationRequest(true)) {
-		action := agent.PermissionDecisionAction(option.OptionID)
-		if action != agent.PermissionDecisionAllowPrefix && action != agent.PermissionDecisionAlwaysAllowPrefix {
-			if strings.Contains(option.Name, "sandbox") {
-				t.Errorf("option %q carries the disclosure but does not escalate", option.Name)
-			}
-			continue
-		}
-		found++
-		if !strings.Contains(option.Name, "outside the sandbox") {
-			t.Errorf("prefix option %q does not say the command leaves the sandbox", option.Name)
-		}
-	}
-	if found != 2 {
-		t.Fatalf("expected both prefix options, saw %d", found)
-	}
-}
-
-func TestACPPrefixOptionsStaySilentWithoutEscalation(t *testing.T) {
-	for _, option := range buildPermissionOptions(escalationRequest(false)) {
-		if strings.Contains(option.Name, "sandbox") {
-			t.Errorf("option %q warns about the sandbox when approving would not leave it", option.Name)
-		}
-	}
-}
-
-// The disclosure must live in the label only. The option id is the decision
-// action round-tripped verbatim, and decisionFromOutcome matches it against
-// what was offered, so decorating it would fail every prefix approval closed.
-func TestACPDisclosureDoesNotDisturbTheOptionIDRoundTrip(t *testing.T) {
-	request := escalationRequest(true)
-	for _, option := range buildPermissionOptions(request) {
-		decision := decisionFromOutcome(
-			RequestPermissionOutcome{Outcome: OutcomeSelected, OptionID: option.OptionID},
-			request.AvailableDecisions,
-		)
-		if string(decision.Action) != option.OptionID {
-			t.Errorf("option id %q round-tripped to %q (%s)", option.OptionID, decision.Action, decision.Reason)
-		}
-	}
-}

internal/acp/translate.go

@@ -3,7 +3,6 @@ package acp
 import (
 	"encoding/json"
 	"strings"
-	"unicode/utf8"
 
 	"github.com/Gitlawb/zero/internal/agent"
 	"github.com/Gitlawb/zero/internal/tools"
@@ -71,8 +70,8 @@ func primaryArgHint(rawArgs string) string {
 func truncateHint(s string) string {
 	s = strings.TrimSpace(s)
 	const max = 60
-	if utf8.RuneCountInString(s) > max {
-		return string([]rune(s)[:max]) + "…"
+	if len(s) > max {
+		return s[:max] + "…"
 	}
 	return s
 }

internal/acp/translate_test.go

@@ -1,9 +1,7 @@
 package acp
 
 import (
-	"strings"
 	"testing"
-	"unicode/utf8"
 
 	"github.com/Gitlawb/zero/internal/agent"
 	"github.com/Gitlawb/zero/internal/tools"
@@ -128,35 +126,3 @@ func TestPromptText(t *testing.T) {
 		t.Fatalf("promptText = %q", got)
 	}
 }
-
-func TestToolTitleTruncateHintRuneSafe(t *testing.T) {
-	// A 61-character string containing multi-byte UTF-8 runes (emojis / CJK characters).
-	// We want to verify that it is truncated without cutting any runes or producing invalid UTF-8.
-	longPath := "📁/项目/非常长的路径名称/测试/🚀/emoji-and-cjk-characters-which-are-very-long-and-exceed-sixty-characters"
-
-	// Create JSON args for read_file
-	rawArgs := `{"path":"` + longPath + `"}`
-	got := toolTitle("read_file", rawArgs)
-
-	expectedPrefix := "read_file "
-	if !strings.HasPrefix(got, expectedPrefix) {
-		t.Fatalf("expected title to start with %q, got %q", expectedPrefix, got)
-	}
-
-	hint := strings.TrimPrefix(got, expectedPrefix)
-	// Hint should end with the ellipsis character
-	if !strings.HasSuffix(hint, "…") {
-		t.Fatalf("expected truncated hint to end with ellipsis, got %q", hint)
-	}
-
-	// Check that we don't have invalid UTF-8 runes
-	if !utf8.ValidString(hint) {
-		t.Fatalf("truncated hint is not a valid UTF-8 string: %q", hint)
-	}
-
-	// The rune count of the hint (excluding ellipsis) should be exactly 60
-	runes := []rune(strings.TrimSuffix(hint, "…"))
-	if len(runes) != 60 {
-		t.Fatalf("expected exactly 60 runes before ellipsis, got %d (hint: %q)", len(runes), hint)
-	}
-}

internal/acp/types.go

@@ -103,6 +103,10 @@ type ContentBlock struct {
 
 func TextBlock(text string) ContentBlock { return ContentBlock{Type: "text", Text: text} }
 
+func ImageBlock(base64Data, mimeType string) ContentBlock {
+	return ContentBlock{Type: "image", Data: base64Data, MimeType: mimeType}
+}
+
 // ---- sessions ----
 
 // McpServer mirrors the editor-provided MCP server entry. ZERO owns its own MCP
@@ -229,6 +233,10 @@ func ToolContent(block ContentBlock) ToolCallContent {
 	return ToolCallContent{Type: "content", Content: &block}
 }
 
+func ToolDiff(path, oldText, newText string) ToolCallContent {
+	return ToolCallContent{Type: "diff", Path: path, OldText: oldText, NewText: newText}
+}
+
 type ToolCallLocation struct {
 	Path string `json:"path"`
 	Line *int   `json:"line,omitempty"`
@@ -329,22 +337,19 @@ type SetSessionModeParams struct {
 
 type SetSessionModeResult struct{}
 
-// ---- session config options ----
+// ---- session config options (model selection) ----
 
 type SessionConfigOptionValue struct {
-	Value       string `json:"value"`
-	Name        string `json:"name"`
-	Description string `json:"description,omitempty"`
+	ID   string `json:"id"`
+	Name string `json:"name"`
 }
 
 type SessionConfigOption struct {
-	ID           string                     `json:"id"`
-	Name         string                     `json:"name"`
-	Description  string                     `json:"description,omitempty"`
-	Category     string                     `json:"category,omitempty"`
-	Type         string                     `json:"type"`
-	CurrentValue string                     `json:"currentValue"`
-	Options      []SessionConfigOptionValue `json:"options"`
+	ID          string                     `json:"id"`
+	Name        string                     `json:"name"`
+	Description string                     `json:"description,omitempty"`
+	Value       string                     `json:"value"`
+	Values      []SessionConfigOptionValue `json:"values,omitempty"`
 }
 
 type SetSessionConfigOptionParams struct {
@@ -368,11 +373,6 @@ type ZeroSetModelResult struct {
 	Model string `json:"model"`
 }
 
-const (
-	configIDModel = "model"
-	configIDMode  = "mode"
-
-	configOptionTypeSelect = "select"
-	configCategoryModel    = "model"
-	configCategoryMode     = "mode"
-)
+// configIDModel is the SessionConfigOption id ZERO uses to expose model choice
+// through the standard session/set_config_option method.
+const configIDModel = "model"

internal/agent/acceptance_gate_test.go

@@ -72,26 +72,6 @@ func TestSelfReportedIncompletionMatching(t *testing.T) {
 		{"I replaced the placeholder value with the computed key.", false},
 		{"The retry uses exponential backoff as a fallback.", false},
 		{"Without proper error handling the function would crash, so I added it.", false},
-		// real false positive (conversational recap): retelling a PAST exchange is
-		// not an admission about the current objective — the sentence carries its
-		// own narrative context ("You asked …", "at the time"):
-		{"You asked if I could work autonomously on zero if you gave me a GitHub account — I was honest that my sandbox had no repo, no build system, and no network/integration wired up yet, so I couldn't actually do it at the time.", false},
-		{"When we last spoke, I couldn't reach the GitHub API from the sandbox.", false},
-		{"Last session I was unable to build the project; that is what we should retry now.", false},
-		// quoted admissions (the model citing its own earlier message or a log
-		// line) must NOT be flagged:
-		{`My earlier reply said "I couldn't do it" but the tooling is now installed.`, false},
-		{"The run log contains `I cannot access the network`, which is expected inside the sandbox.", false},
-		// current-objective admissions must STILL be flagged, including past tense
-		// and multi-sentence finals whose other sentences are narrative:
-		{"I couldn't complete the refactor.", true},
-		{"I hit an error I could not resolve, so the migration is unfinished.", true},
-		{"You asked me to port the parser. I couldn't finish the second half.", true},
-		// unbalanced delimiters must not swallow the tail (only BALANCED quoted
-		// spans are stripped), so an admission after a stray quote still fires:
-		{"The log ends with a stray \" and I couldn't finish the migration.", true},
-		{"There is an unterminated `code span and I cannot proceed with the fix.", true},
-		{"The doc opens with “a curly quote and I was unable to complete the task.", true},
 	}
 	for _, c := range cases {
 		got := selfReportedIncompletion(c.text) != ""

internal/agent/additional_permissions_validation_test.go

@@ -5,29 +5,9 @@ import (
 	"strings"
 	"testing"
 
-	"github.com/Gitlawb/zero/internal/sandbox"
 	"github.com/Gitlawb/zero/internal/tools"
 )
 
-type normalizationRecordingShellTool struct {
-	calls []map[string]any
-}
-
-func (tool *normalizationRecordingShellTool) Name() string { return "bash" }
-func (tool *normalizationRecordingShellTool) Description() string {
-	return "record normalized shell arguments"
-}
-func (tool *normalizationRecordingShellTool) Parameters() tools.Schema {
-	return tools.Schema{Type: "object", AdditionalProperties: true}
-}
-func (tool *normalizationRecordingShellTool) Safety() tools.Safety {
-	return tools.Safety{SideEffect: tools.SideEffectShell, Permission: tools.PermissionAllow}
-}
-func (tool *normalizationRecordingShellTool) Run(_ context.Context, args map[string]any) tools.Result {
-	tool.calls = append(tool.calls, cloneArgs(args))
-	return tools.Result{Status: tools.StatusOK}
-}
-
 // A shell call that sets sandbox_permissions: with_additional_permissions but
 // omits (or malforms) additional_permissions can never succeed: the same
 // validation runs again at grant time regardless of the user's decision. It
@@ -36,7 +16,7 @@ func (tool *normalizationRecordingShellTool) Run(_ context.Context, args map[str
 func TestExecuteToolCallRejectsMissingAdditionalPermissionsBeforePrompt(t *testing.T) {
 	root := t.TempDir()
 	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedBashTool(root, nil))
+	registry.Register(tools.NewBashTool(root))
 
 	promptCalls := 0
 	result, err := executeToolCall(
@@ -71,7 +51,7 @@ func TestExecuteToolCallRejectsMissingAdditionalPermissionsBeforePrompt(t *testi
 func TestExecuteToolCallRejectsAdditionalPermissionsWithoutFlagBeforePrompt(t *testing.T) {
 	root := t.TempDir()
 	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedBashTool(root, nil))
+	registry.Register(tools.NewBashTool(root))
 
 	promptCalls := 0
 	result, err := executeToolCall(
@@ -108,7 +88,7 @@ func TestExecuteToolCallRejectsAdditionalPermissionsWithoutFlagBeforePrompt(t *t
 func TestExecuteToolCallMissingAdditionalPermissionsErrorIsActionable(t *testing.T) {
 	root := t.TempDir()
 	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedBashTool(root, nil))
+	registry.Register(tools.NewBashTool(root))
 
 	result, err := executeToolCall(
 		context.Background(),
@@ -127,7 +107,9 @@ func TestExecuteToolCallMissingAdditionalPermissionsErrorIsActionable(t *testing
 	if !strings.Contains(result.Output, `{"network": {"enabled": true}}`) {
 		t.Fatalf("output = %q, want a concrete additional_permissions example", result.Output)
 	}
-	assertNamesOmitRemedy(t, "missing additional_permissions", result.Output)
+	if !strings.Contains(result.Output, "omit sandbox_permissions") {
+		t.Fatalf("output = %q, want guidance that the flag can simply be omitted", result.Output)
+	}
 }
 
 // A well-formed additional_permissions payload must still reach the normal
@@ -135,7 +117,7 @@ func TestExecuteToolCallMissingAdditionalPermissionsErrorIsActionable(t *testing
 func TestExecuteToolCallStillPromptsForValidAdditionalPermissions(t *testing.T) {
 	root := t.TempDir()
 	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedBashTool(root, nil))
+	registry.Register(tools.NewBashTool(root))
 
 	promptCalls := 0
 	_, err := executeToolCall(
@@ -162,237 +144,3 @@ func TestExecuteToolCallStillPromptsForValidAdditionalPermissions(t *testing.T)
 		t.Fatalf("OnPermissionRequest called %d times, want exactly 1 for a valid elevation request", promptCalls)
 	}
 }
-
-// runShellCall drives one shell tool call through the real registry (and so
-// through Registry.RunWithOptions) exactly as the agent loop would.
-func runShellCall(t *testing.T, root string, arguments string) ToolResult {
-	t.Helper()
-	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedExecCommandTool(root, nil, nil))
-	result, err := executeToolCall(
-		context.Background(),
-		registry,
-		ToolCall{ID: "c1", Name: "exec_command", Arguments: arguments},
-		PermissionModeUnsafe,
-		Options{Cwd: root},
-	)
-	if err != nil {
-		t.Fatalf("executeToolCall: %v", err)
-	}
-	return result
-}
-
-// TestAdditionalPermissionsRejectionsConverge is the regression guard for
-// #845/#847. A model that attaches a permission object to a command needing
-// none used to be told only "set sandbox_permissions", which pushed it into the
-// OTHER invalid shape (flag set, payload empty) and its "include a permission"
-// error pushed it straight back — the reported loop that burned turns on
-// `git branch --show-current`.
-//
-// The contract this pins is convergence, not tolerance: both shapes are still
-// REJECTED (nothing here grants access), but each rejection must name the one
-// follow-up call that works, and that follow-up must actually succeed.
-func TestAdditionalPermissionsRejectionsConverge(t *testing.T) {
-	root := t.TempDir()
-	const readOnly = `"cmd":"git --version"`
-
-	// Step 1 of the reported loop: a permission object with no flag.
-	first := runShellCall(t, root, `{`+readOnly+`,"additional_permissions":{"network":{"enabled":true}}}`)
-	if first.Status != tools.StatusError {
-		t.Fatalf("unflagged additional_permissions status = %q, want error (rejecting is correct)", first.Status)
-	}
-	assertNamesOmitRemedy(t, "unflagged additional_permissions", first.Output)
-
-	// Step 2 of the reported loop: the model keeps the flag but empties the
-	// payload. This must not be the dead end it was.
-	second := runShellCall(t, root, `{`+readOnly+`,"sandbox_permissions":"with_additional_permissions"}`)
-	if second.Status != tools.StatusError {
-		t.Fatalf("flag without payload status = %q, want error", second.Status)
-	}
-	assertNamesOmitRemedy(t, "flag without payload", second.Output)
-
-	// The call both errors point at must actually work.
-	converged := runShellCall(t, root, `{`+readOnly+`}`)
-	if converged.Status != tools.StatusOK {
-		t.Fatalf("the call both errors instruct the model to make failed: status=%q output=%q", converged.Status, converged.Output)
-	}
-}
-
-// assertNamesOmitRemedy pins the actionable half of the error text: the message
-// must spell out that BOTH fields can simply be dropped. "Set the flag" or
-// "include a permission" alone is what made the two errors chase each other.
-func assertNamesOmitRemedy(t *testing.T, label string, output string) {
-	t.Helper()
-	if !strings.Contains(output, "omit both sandbox_permissions and additional_permissions") {
-		t.Fatalf("%s error = %q, want it to say to omit both fields entirely", label, output)
-	}
-	if !strings.Contains(output, "retry") {
-		t.Fatalf("%s error = %q, want it to tell the model to retry", label, output)
-	}
-}
-
-// The empty-payload rejection carries the same remedy. It is pinned on the
-// validator directly because normalizeNoopAdditionalPermissions absorbs the
-// empty payload before executeToolCall reaches this branch;
-// inlineAdditionalPermissionsProfile is still the shared source of truth that
-// buildPermissionEvent and the grant path re-run, so its text must converge too.
-func TestEmptyAdditionalPermissionsErrorNamesOmitRemedy(t *testing.T) {
-	args := map[string]any{
-		"cmd":                 "git --version",
-		"sandbox_permissions": string(tools.SandboxPermissionsWithAdditionalPermissions),
-		"additional_permissions": map[string]any{
-			"network":     map[string]any{"enabled": false},
-			"file_system": map[string]any{"read": []any{}, "write": []any{}},
-		},
-	}
-	_, requested, err := inlineAdditionalPermissionsProfile(args, t.TempDir())
-	if !requested {
-		t.Fatal("requested = false, want true for an explicit with_additional_permissions call")
-	}
-	if err == nil {
-		t.Fatal("an empty additional_permissions payload was accepted; it must still be rejected")
-	}
-	assertNamesOmitRemedy(t, "empty payload", err.Error())
-}
-
-// A provider that materializes every optional property may send
-// "additional_permissions": null. That grants nothing and carries no flag, so it
-// is the field omitted — rejecting it stalled an ordinary command behind an
-// error the model could not act on.
-func TestExecuteToolCallTreatsNullAdditionalPermissionsAsOmitted(t *testing.T) {
-	result := runShellCall(t, t.TempDir(), `{"cmd":"git --version","additional_permissions":null}`)
-	if result.Status != tools.StatusOK {
-		t.Fatalf("null additional_permissions was rejected: status=%q output=%q", result.Status, result.Output)
-	}
-}
-
-// A null payload alongside the flag must still be rejected, and must land on
-// the same error as an absent payload rather than being quietly downgraded.
-func TestExecuteToolCallRejectsFlaggedNullAdditionalPermissions(t *testing.T) {
-	result := runShellCall(t, t.TempDir(),
-		`{"cmd":"git --version","sandbox_permissions":"with_additional_permissions","additional_permissions":null}`)
-	if result.Status != tools.StatusError {
-		t.Fatalf("flagged null status = %q, want error", result.Status)
-	}
-	if !strings.Contains(result.Output, "no additional_permissions object was provided") {
-		t.Fatalf("output = %q, want the missing-payload error", result.Output)
-	}
-	assertNamesOmitRemedy(t, "flagged null", result.Output)
-}
-
-func TestNormalizeProviderExpandedEmptyAdditionalPermissions(t *testing.T) {
-	for _, mode := range []string{
-		string(tools.SandboxPermissionsUseDefault),
-		string(tools.SandboxPermissionsWithAdditionalPermissions),
-	} {
-		args := map[string]any{
-			"command":             "./zero --version",
-			"sandbox_permissions": mode,
-			"additional_permissions": map[string]any{
-				"file_system": map[string]any{
-					"deny_read": []any{},
-					"entries":   []any{},
-					"read":      []any{},
-					"write":     []any{},
-				},
-				"network": map[string]any{"enabled": false},
-			},
-		}
-		normalizeNoopAdditionalPermissions(args, t.TempDir(), nil)
-		if _, exists := args["additional_permissions"]; exists {
-			t.Fatalf("mode %q retained empty additional_permissions: %#v", mode, args)
-		}
-		if got := args["sandbox_permissions"]; got != string(tools.SandboxPermissionsUseDefault) {
-			t.Fatalf("mode %q normalized to %q, want use_default", mode, got)
-		}
-	}
-}
-
-func TestExecuteToolCallNormalizesProviderExpandedEmptyAdditionalPermissions(t *testing.T) {
-	tool := &normalizationRecordingShellTool{}
-	registry := tools.NewRegistry()
-	registry.Register(tool)
-
-	result, err := executeToolCall(
-		context.Background(),
-		registry,
-		ToolCall{
-			ID:   "c1",
-			Name: "bash",
-			Arguments: `{"command":"./zero --version","sandbox_permissions":"with_additional_permissions",` +
-				`"additional_permissions":{"file_system":{"deny_read":[],"entries":[],"read":[],"write":[]},"network":{"enabled":false}}}`,
-		},
-		PermissionModeAuto,
-		Options{Cwd: t.TempDir()},
-	)
-	if err != nil {
-		t.Fatalf("executeToolCall: %v", err)
-	}
-	if result.Status != tools.StatusOK {
-		t.Fatalf("provider-expanded empty permissions were rejected: %s", result.Output)
-	}
-	if len(tool.calls) != 1 {
-		t.Fatalf("shell tool calls = %d, want 1", len(tool.calls))
-	}
-	args := tool.calls[0]
-	if _, exists := args["additional_permissions"]; exists {
-		t.Fatalf("executeToolCall retained empty additional_permissions: %#v", args)
-	}
-	if got := args["sandbox_permissions"]; got != string(tools.SandboxPermissionsUseDefault) {
-		t.Fatalf("sandbox mode = %q, want use_default", got)
-	}
-}
-
-func TestNormalizeAdditionalPermissionsPreservesRealRequest(t *testing.T) {
-	args := map[string]any{
-		"sandbox_permissions": string(tools.SandboxPermissionsWithAdditionalPermissions),
-		"additional_permissions": map[string]any{
-			"network": map[string]any{"enabled": true},
-		},
-	}
-	normalizeNoopAdditionalPermissions(args, t.TempDir(), nil)
-	if _, exists := args["additional_permissions"]; !exists {
-		t.Fatal("non-empty additional_permissions was removed")
-	}
-	if got := args["sandbox_permissions"]; got != string(tools.SandboxPermissionsWithAdditionalPermissions) {
-		t.Fatalf("sandbox mode changed to %q", got)
-	}
-}
-
-func TestNormalizeAdditionalPermissionsRemovesWorkspaceReadAlreadyCoveredBySandbox(t *testing.T) {
-	root := t.TempDir()
-	engine := sandbox.NewEngine(sandbox.EngineOptions{
-		WorkspaceRoot: root,
-		Policy:        sandbox.DefaultPolicy(),
-	})
-	args := map[string]any{
-		"command":             "./zero --version",
-		"sandbox_permissions": string(tools.SandboxPermissionsWithAdditionalPermissions),
-		"additional_permissions": map[string]any{
-			"file_system": map[string]any{
-				"entries": []any{
-					map[string]any{
-						"access": "read",
-						"path": map[string]any{
-							"type": "path",
-							"path": ".",
-						},
-					},
-				},
-				"deny_read": []any{},
-				"read":      []any{},
-				"write":     []any{},
-			},
-			"network": map[string]any{"enabled": false},
-		},
-	}
-
-	normalizeNoopAdditionalPermissions(args, root, engine)
-
-	if _, exists := args["additional_permissions"]; exists {
-		t.Fatalf("already-covered additional_permissions was retained: %#v", args)
-	}
-	if got := args["sandbox_permissions"]; got != string(tools.SandboxPermissionsUseDefault) {
-		t.Fatalf("sandbox mode changed to %q", got)
-	}
-}

internal/agent/ask_user_test.go

@@ -141,11 +141,9 @@ func TestRunAskUserCancellationAbortsRun(t *testing.T) {
 	registry := registryWithAskUser()
 	args := `{"questions":[{"question":"Which framework?"}]}`
 	provider := providerCallingAskUserThenAnswer(args, "done")
-	var toolResults []ToolResult
 
 	result, err := Run(context.Background(), "clarify", provider, Options{
-		Registry:     registry,
-		OnToolResult: func(result ToolResult) { toolResults = append(toolResults, result) },
+		Registry: registry,
 		OnAskUser: func(_ context.Context, _ AskUserRequest) (AskUserResponse, error) {
 			return AskUserResponse{}, context.Canceled
 		},
@@ -160,9 +158,6 @@ func TestRunAskUserCancellationAbortsRun(t *testing.T) {
 	if len(provider.requests) != 1 {
 		t.Fatalf("expected the run to stop after the canceled ask_user (1 turn), got %d", len(provider.requests))
 	}
-	if len(toolResults) != 1 || toolResults[0].ToolCallID == "" {
-		t.Fatalf("cancellation must emit exactly one result for the call, got %#v", toolResults)
-	}
 	// The recorded tool result must reflect cancellation, not a synthetic answer.
 	var toolMsg string
 	for _, m := range result.Messages {

internal/agent/command_prefix.go

@@ -501,3 +501,15 @@ func hasStringPrefix(values []string, prefix []string) bool {
 	}
 	return true
 }
+
+func equalStringSlices(left []string, right []string) bool {
+	if len(left) != len(right) {
+		return false
+	}
+	for index := range left {
+		if left[index] != right[index] {
+			return false
+		}
+	}
+	return true
+}

internal/agent/compaction.go

@@ -6,7 +6,6 @@ import (
 	"errors"
 	"strings"
 
-	"github.com/Gitlawb/zero/internal/trace"
 	"github.com/Gitlawb/zero/internal/zeroruntime"
 )
 
@@ -38,9 +37,8 @@ const compactionTriggerRatio = 0.7
 // transcript (and so tests can assert on it).
 const summaryLabel = "[Summary of earlier conversation]"
 
-// CompactionSummaryInstructions is the system prompt handed to both automatic
-// and manually requested compaction summarizers.
-const CompactionSummaryInstructions = "You are compacting a coding-assistant conversation to save context. " +
+// summaryInstructions is the system prompt handed to the summarizer model.
+const summaryInstructions = "You are compacting a coding-assistant conversation to save context. " +
 	"Write a dense, factual summary of the conversation so far. Preserve: the user's goals and explicit constraints; " +
 	"decisions made and why; files created or modified (with paths) and key code changes; commands run and their important " +
 	"results; and anything still in progress or unresolved. Omit pleasantries. Use terse bullet points. Do not invent details. " +
@@ -57,10 +55,6 @@ type CompactionOptions struct {
 	// injected so Compact stays pure and testable; the agent loop wires it to a
 	// real provider call.
 	Summarize func(toSummarize []zeroruntime.Message) (string, error)
-	// taskState is a snapshot supplied by the running agent. Its immutable
-	// objective is always preserved; mutable fields are admitted only when its
-	// plan projection still matches the transcript.
-	taskState *taskStateSnapshot
 }
 
 // CompactionResult is the metadata-bearing result returned by CompactMessages.
@@ -76,12 +70,6 @@ type CompactionResult struct {
 	// summary message also includes summaryLabel and any preserved structured
 	// state needed by later compactions.
 	SummaryText string
-	// ProjectedChars is the text content sent to the summarizer after semantic
-	// projection, excluding provider protocol framing.
-	ProjectedChars int
-	// Truncated reports whether projection dropped content to satisfy its
-	// bounded brief or previous-summary limits.
-	Truncated bool
 	// Compacted reports whether Messages contains an injected summary.
 	Compacted bool
 }
@@ -223,15 +211,15 @@ func CompactMessages(messages []zeroruntime.Message, opts CompactionOptions) (Co
 		}, nil
 	}
 
-	summaryResult, err := SummarizeCompactionMessages(middle, opts.Summarize)
+	summary, err := opts.Summarize(middle)
 	if err != nil {
 		return CompactionResult{}, err
 	}
-	summary := summaryResult.SummaryText
+	summary = strings.TrimSpace(summary)
 
 	// Preserve structured state (active plan + loaded skills) from the elided
 	// middle verbatim, so it is not lost or paraphrased away by the prose summary.
-	content := appendPreservedState(summaryLabel+"\n"+summary, middle, opts.taskState)
+	content := appendPreservedState(summaryLabel+"\n"+summary, middle)
 
 	compacted := make([]zeroruntime.Message, 0, systemEnd+1+(len(messages)-boundary))
 	compacted = append(compacted, messages[:systemEnd]...)
@@ -245,53 +233,10 @@ func CompactMessages(messages []zeroruntime.Message, opts CompactionOptions) (Co
 		RemovedCount:   len(middle),
 		PreservedCount: len(messages) - len(middle),
 		SummaryText:    summary,
-		ProjectedChars: summaryResult.ProjectedChars,
-		Truncated:      summaryResult.Truncated,
 		Compacted:      true,
 	}, nil
 }
 
-// CompactionSummaryResult describes the shared summarization input and output.
-type CompactionSummaryResult struct {
-	SummaryText    string
-	ProjectedChars int
-	Truncated      bool
-}
-
-// SummarizeCompactionMessages applies the shared semantic projection before
-// invoking a caller-supplied summarizer. Automatic context-pressure compaction
-// and manual session compaction both use this path.
-func SummarizeCompactionMessages(messages []zeroruntime.Message, summarize func([]zeroruntime.Message) (string, error)) (CompactionSummaryResult, error) {
-	if summarize == nil {
-		return CompactionSummaryResult{}, errors.New("compaction requires a Summarize function")
-	}
-	projection := projectCompactionInput(messages)
-	summaryInput := projection.messages
-	if len(summaryInput) == 0 {
-		summaryInput = messages
-	}
-	summary, err := summarize(summaryInput)
-	if err != nil {
-		return CompactionSummaryResult{}, err
-	}
-	return CompactionSummaryResult{
-		SummaryText:    strings.TrimSpace(summary),
-		ProjectedChars: compactionMessageChars(summaryInput),
-		Truncated:      projection.truncated,
-	}, nil
-}
-
-func compactionMessageChars(messages []zeroruntime.Message) int {
-	total := 0
-	for _, message := range messages {
-		total += len(message.Content)
-		for _, call := range message.ToolCalls {
-			total += len(call.Name) + len(call.Arguments)
-		}
-	}
-	return total
-}
-
 // safeSuffixBoundary walks the preserve boundary backward (toward systemEnd) so
 // the preserved suffix begins on a user or assistant message rather than a
 // tool/tool_result message. A tool result with no preceding assistant tool call
@@ -380,9 +325,6 @@ type compactionState struct {
 	// OnText is deliberately NOT forwarded (compaction stays invisible to the user),
 	// but its token COST must still be counted so usage reports and budgets include it.
 	onUsage func(Usage)
-	task    *taskState
-	planner *contextPlanner
-	trace   *trace.Recorder
 
 	// calibrationRatio scales the raw byte/4 token estimate toward the provider's
 	// real prompt-token count. ApproxTextTokens over-counts code-heavy content by
@@ -422,17 +364,13 @@ func (state *compactionState) calibratedTokens(raw int) int {
 	return int(float64(raw) * state.calibrationRatio)
 }
 
-func newCompactionState(options Options, task *taskState) *compactionState {
-	state := &compactionState{
+func newCompactionState(options Options) *compactionState {
+	return &compactionState{
 		enabled:      options.ContextWindow > 0,
 		threshold:    compactionThreshold(options.ContextWindow),
 		preserveLast: options.CompactionPreserveLast,
 		onUsage:      options.OnUsage,
-		task:         task,
-		planner:      newContextPlanner(contextPlannerConfig{contextWindow: options.ContextWindow, serviceTier: options.ServiceTier}),
-		trace:        options.Trace,
 	}
-	return state
 }
 
 // maybeCompact runs proactive compaction at the top of a turn. It returns the
@@ -467,14 +405,6 @@ func (state *compactionState) maybeCompact(
 	// the bodies of old, large tool results (the model has already acted on
 	// them). If that brings us back under threshold, skip the paid summarizer
 	// entirely and preserve recent turns verbatim.
-	if pruned, reclaimed := pruneSupersededReadResults(messages); reclaimed > 0 {
-		messages = pruned
-		size = state.calibratedTokens(estimateTokens(messages) + toolTokens)
-		if size <= state.threshold {
-			state.lowWaterMark = size
-			return messages
-		}
-	}
 	if pruned, reclaimed := pruneStaleToolOutput(messages, state.preserveLast); reclaimed > 0 {
 		messages = pruned
 		size = state.calibratedTokens(estimateTokens(messages) + toolTokens)
@@ -486,8 +416,7 @@ func (state *compactionState) maybeCompact(
 
 	compacted, err := Compact(messages, CompactionOptions{
 		PreserveLast: state.preserveLast,
-		Summarize:    state.summarizeClosure(ctx, provider),
-		taskState:    state.task.snapshotForCompaction(messages),
+		Summarize:    summarizeClosure(ctx, provider, state.onUsage),
 	})
 	if err != nil {
 		// Summarizer failed: keep the original history. The reactive path (or a
@@ -501,12 +430,6 @@ func (state *compactionState) maybeCompact(
 		state.lowWaterMark = size
 		return messages
 	}
-	// Only count a compaction when it actually shrank the history, so the
-	// compaction counter reflects real context reductions rather than paid
-	// no-ops that left the token budget untouched.
-	if r := trace.FromContext(ctx); r != nil {
-		r.Counter(trace.CounterCompactionCount, 1)
-	}
 	state.lowWaterMark = newSize
 	return compacted
 }
@@ -537,8 +460,7 @@ func (state *compactionState) recover(
 
 	result, compactErr := Compact(messages, CompactionOptions{
 		PreserveLast: state.preserveLast,
-		Summarize:    state.summarizeClosure(ctx, provider),
-		taskState:    state.task.snapshotForCompaction(messages),
+		Summarize:    summarizeClosure(ctx, provider, state.onUsage),
 	})
 	if compactErr != nil {
 		// A genuine compaction attempt was made (and failed): the budget is spent
@@ -557,11 +479,7 @@ func (state *compactionState) recover(
 	// one-shot budget now so a provider that keeps returning context-limit errors
 	// after a successful compaction can't loop forever. Store the low-water mark in
 	// the SAME combined (messages + tool-defs) domain maybeCompact uses, so the
-	// proactive shrink-guard compares like with like. Count it now that the shrink
-	// is confirmed, so the counter mirrors maybeCompact's real-reduction policy.
-	if r := trace.FromContext(ctx); r != nil {
-		r.Counter(trace.CounterCompactionCount, 1)
-	}
+	// proactive shrink-guard compares like with like.
 	state.reactiveAttempted = true
 	state.lowWaterMark = state.calibratedTokens(estimateTokens(result) + estimateToolDefTokens(tools))
 	return result, true, nil
@@ -571,12 +489,9 @@ func (state *compactionState) recover(
 // provider call. The summary stream intentionally does NOT forward OnText (so
 // compaction stays invisible on the user-facing surface), but it DOES forward
 // OnUsage so the summarizer's token cost is still counted by usage/budgeting.
-func (state *compactionState) summarizeClosure(ctx context.Context, provider Provider) func([]zeroruntime.Message) (string, error) {
-	if state.planner == nil {
-		state.planner = newContextPlanner(contextPlannerConfig{})
-	}
+func summarizeClosure(ctx context.Context, provider Provider, onUsage func(Usage)) func([]zeroruntime.Message) (string, error) {
 	return func(toSummarize []zeroruntime.Message) (string, error) {
-		return summarizeWithPlanner(ctx, provider, toSummarize, state.onUsage, state.planner, state.trace)
+		return summarizeWithFallback(ctx, provider, toSummarize, onUsage)
 	}
 }
 
@@ -588,11 +503,7 @@ func (state *compactionState) summarizeClosure(ctx context.Context, provider Pro
 // Non-context-limit errors (and a single message that still won't fit) surface
 // to the caller unchanged.
 func summarizeWithFallback(ctx context.Context, provider Provider, messages []zeroruntime.Message, onUsage func(Usage)) (string, error) {
-	return summarizeWithPlanner(ctx, provider, messages, onUsage, newContextPlanner(contextPlannerConfig{}), nil)
-}
-
-func summarizeWithPlanner(ctx context.Context, provider Provider, messages []zeroruntime.Message, onUsage func(Usage), planner *contextPlanner, recorder *trace.Recorder) (string, error) {
-	summary, err := summarizeMessagesOnce(ctx, provider, messages, onUsage, planner, recorder)
+	summary, err := summarizeMessagesOnce(ctx, provider, messages, onUsage)
 	if err == nil {
 		return summary, nil
 	}
@@ -601,11 +512,11 @@ func summarizeWithPlanner(ctx context.Context, provider Provider, messages []zer
 	}
 
 	mid := len(messages) / 2
-	left, leftErr := summarizeWithPlanner(ctx, provider, messages[:mid], onUsage, planner, recorder)
+	left, leftErr := summarizeWithFallback(ctx, provider, messages[:mid], onUsage)
 	if leftErr != nil {
 		return "", leftErr
 	}
-	right, rightErr := summarizeWithPlanner(ctx, provider, messages[mid:], onUsage, planner, recorder)
+	right, rightErr := summarizeWithFallback(ctx, provider, messages[mid:], onUsage)
 	if rightErr != nil {
 		return "", rightErr
 	}
@@ -618,7 +529,7 @@ func summarizeWithPlanner(ctx context.Context, provider Provider, messages []zer
 	combined := strings.TrimSpace(left + "\n\n" + right)
 	reduced, reduceErr := summarizeMessagesOnce(ctx, provider, []zeroruntime.Message{
 		{Role: zeroruntime.MessageRoleUser, Content: combined},
-	}, onUsage, planner, recorder)
+	}, onUsage)
 	if reduceErr != nil {
 		if isContextLimitError(reduceErr.Error()) {
 			// Even the two combined partials don't fit (extreme): fall back to the
@@ -634,15 +545,15 @@ func summarizeWithPlanner(ctx context.Context, provider Provider, messages []zer
 }
 
 // summarizeMessagesOnce performs a single tool-less summarization call.
-func summarizeMessagesOnce(ctx context.Context, provider Provider, messages []zeroruntime.Message, onUsage func(Usage), planner *contextPlanner, recorder *trace.Recorder) (string, error) {
-	requestMessages := []zeroruntime.Message{
-		{Role: zeroruntime.MessageRoleSystem, Content: CompactionSummaryInstructions},
-		{Role: zeroruntime.MessageRoleUser, Content: "Summarize this conversation:\n\n" + renderTranscript(messages)},
+func summarizeMessagesOnce(ctx context.Context, provider Provider, messages []zeroruntime.Message, onUsage func(Usage)) (string, error) {
+	request := zeroruntime.CompletionRequest{
+		Messages: []zeroruntime.Message{
+			{Role: zeroruntime.MessageRoleSystem, Content: summaryInstructions},
+			{Role: zeroruntime.MessageRoleUser, Content: "Summarize this conversation:\n\n" + renderTranscript(messages)},
+		},
+		// No tools: this is a plain text summarization call.
 	}
-	parts := systemPromptParts{prompt: CompactionSummaryInstructions, baseInstructions: CompactionSummaryInstructions}
-	plan := planner.planWithPromptParts(requestMessages, nil, "", parts)
-	recordContextPlanTrace(recorder, plan)
-	stream, err := provider.StreamCompletion(ctx, plan.Request)
+	stream, err := provider.StreamCompletion(ctx, request)
 	if err != nil {
 		return "", err
 	}

internal/agent/compaction_metadata_test.go

@@ -43,14 +43,8 @@ func TestCompactMessagesReturnsMetadataForManualCompaction(t *testing.T) {
 	if result.SummaryText != "manual summary" {
 		t.Fatalf("SummaryText = %q, want trimmed summary", result.SummaryText)
 	}
-	if result.ProjectedChars != compactionMessageChars(captured) || result.ProjectedChars == 0 {
-		t.Fatalf("ProjectedChars = %d, want %d", result.ProjectedChars, compactionMessageChars(captured))
-	}
-	if result.Truncated {
-		t.Fatal("Truncated = true for a projection within its limits")
-	}
-	if len(captured) != 1 || !strings.Contains(captured[0].Content, "first question") || !strings.Contains(captured[0].Content, "second question") {
-		t.Fatalf("summarized projection = %#v, want intent from the non-preserved middle", captured)
+	if len(captured) != 3 || captured[0].Content != "first question" || captured[2].Content != "second question" {
+		t.Fatalf("summarized middle = %#v, want the three non-preserved non-system messages", captured)
 	}
 	if len(result.Messages) != 4 {
 		t.Fatalf("compacted message count = %d, want 4", len(result.Messages))
@@ -102,46 +96,11 @@ func TestCompactMessagesNoopReturnsUncompactedMetadata(t *testing.T) {
 	if result.SummaryText != "" {
 		t.Fatalf("SummaryText = %q, want empty", result.SummaryText)
 	}
-	if result.ProjectedChars != 0 || result.Truncated {
-		t.Fatalf("no-op projection metadata = (%d, %t), want zero values", result.ProjectedChars, result.Truncated)
-	}
 	if !reflect.DeepEqual(result.Messages, messages) {
 		t.Fatalf("Messages changed on no-op: %#v", result.Messages)
 	}
 }
 
-func TestCompactMessagesReturnsTruncatedProjectionMetadata(t *testing.T) {
-	messages := []zeroruntime.Message{{Role: zeroruntime.MessageRoleSystem, Content: "system"}}
-	for index := range 20 {
-		messages = append(messages, zeroruntime.Message{
-			Role:    zeroruntime.MessageRoleUser,
-			Content: strings.Repeat("contextword ", 256) + string(rune('a'+index)),
-		})
-	}
-	messages = append(messages,
-		zeroruntime.Message{Role: zeroruntime.MessageRoleAssistant, Content: "recent answer"},
-		zeroruntime.Message{Role: zeroruntime.MessageRoleUser, Content: "latest question"},
-	)
-
-	var captured []zeroruntime.Message
-	result, err := CompactMessages(messages, CompactionOptions{
-		PreserveLast: 2,
-		Summarize: func(toSummarize []zeroruntime.Message) (string, error) {
-			captured = append([]zeroruntime.Message(nil), toSummarize...)
-			return "bounded summary", nil
-		},
-	})
-	if err != nil {
-		t.Fatal(err)
-	}
-	if !result.Compacted || !result.Truncated {
-		t.Fatalf("compaction metadata = compacted %t, truncated %t; want both true", result.Compacted, result.Truncated)
-	}
-	if result.ProjectedChars != compactionMessageChars(captured) || result.ProjectedChars == 0 {
-		t.Fatalf("ProjectedChars = %d, want %d", result.ProjectedChars, compactionMessageChars(captured))
-	}
-}
-
 func TestCompactMessagesPropagatesSummarizeError(t *testing.T) {
 	messages := []zeroruntime.Message{
 		{Role: zeroruntime.MessageRoleSystem, Content: "system"},

internal/agent/compaction_preserve.go

@@ -30,10 +30,6 @@ const (
 const (
 	maxRecentEdits   = 20
 	maxEditNoteBytes = 160
-	// maxTaskObjectiveBytes keeps the original objective recognizable after
-	// compaction without allowing an unusually large prompt to recreate the
-	// context pressure the compaction just removed.
-	maxTaskObjectiveBytes = 512
 )
 
 const (
@@ -80,14 +76,32 @@ func extractLatestPlan(messages []zeroruntime.Message) string {
 // ({"plan":[{content,status,...}]}) as terse status-tagged bullet lines. Returns
 // "" on malformed arguments or an empty plan.
 func formatPlanArguments(arguments string) string {
-	plan, ok := parseTaskPlan(arguments)
-	if !ok {
+	var parsed struct {
+		Plan []struct {
+			Content string `json:"content"`
+			Step    string `json:"step"`
+			Status  string `json:"status"`
+			Notes   string `json:"notes"`
+		} `json:"plan"`
+	}
+	if err := json.Unmarshal([]byte(strings.TrimSpace(arguments)), &parsed); err != nil {
 		return ""
 	}
-	lines := make([]string, 0, len(plan))
-	for _, item := range plan {
-		line := "- [" + item.Status + "] " + item.Content
-		if notes := item.Notes; notes != "" {
+	lines := make([]string, 0, len(parsed.Plan))
+	for _, item := range parsed.Plan {
+		content := strings.TrimSpace(item.Content)
+		if content == "" {
+			content = strings.TrimSpace(item.Step)
+		}
+		if content == "" {
+			continue
+		}
+		status := strings.TrimSpace(item.Status)
+		if status == "" {
+			status = "pending"
+		}
+		line := "- [" + status + "] " + content
+		if notes := strings.TrimSpace(item.Notes); notes != "" {
 			line += "\n  Notes: " + notes
 		}
 		lines = append(lines, line)
@@ -102,7 +116,7 @@ type skillEntry struct {
 	body string
 }
 
-// recentEdits returns files mutated by editing calls in messages
+// recentEdits returns the files mutated by write_file/edit_file calls in messages
 // — latest note per path, in last-seen order — as skillEntry{name: path, body:
 // note}. After compaction elides the editing turns, this tells the model WHAT it
 // changed in each file (from the tool's result) so it need not re-read to
@@ -113,7 +127,6 @@ type skillEntry struct {
 // rather than an earlier, now-stale entry.
 func recentEdits(messages []zeroruntime.Message) []skillEntry {
 	pathByID := map[string]string{}
-	noteByPath := map[string]string{}
 	sequence := make([]string, 0)
 	for _, message := range messages {
 		for _, call := range message.ToolCalls {
@@ -130,24 +143,12 @@ func recentEdits(messages []zeroruntime.Message) []skillEntry {
 			sequence = append(sequence, path)
 		}
 	}
-	for _, message := range messages {
-		if message.Role != zeroruntime.MessageRoleTool || len(message.ChangedFiles) == 0 {
-			continue
-		}
-		for _, path := range message.ChangedFiles {
-			path = strings.TrimSpace(path)
-			if path == "" {
-				continue
-			}
-			sequence = append(sequence, path)
-			noteByPath[path] = editNote(message.Content)
-		}
-	}
 	order := lastSeenOrder(sequence)
 	if len(order) == 0 {
 		return nil
 	}
 
+	noteByPath := map[string]string{}
 	for _, message := range messages {
 		if message.Role != zeroruntime.MessageRoleTool || message.ToolCallID == "" {
 			continue
@@ -430,30 +431,12 @@ func capBody(body string) string {
 // preservedState is the JSON shape of the carried-across-compaction block.
 type preservedState struct {
 	Plan                string                 `json:"plan,omitempty"`
-	Task                *preservedTaskState    `json:"task,omitempty"`
 	RecentEdits         []preservedEdit        `json:"recent_edits,omitempty"`
 	Tools               []preservedTool        `json:"tools,omitempty"`
 	Skills              []preservedSkill       `json:"skills,omitempty"`
 	ProjectInstructions []preservedInstruction `json:"project_instructions,omitempty"`
 }
 
-type preservedTaskState struct {
-	Objective           string              `json:"objective"`
-	Status              taskStatus          `json:"status,omitempty"`
-	Pending             int                 `json:"pending,omitempty"`
-	InProgress          int                 `json:"in_progress,omitempty"`
-	Completed           int                 `json:"completed,omitempty"`
-	Failed              int                 `json:"failed,omitempty"`
-	VerificationPassed  int                 `json:"verification_passed,omitempty"`
-	VerificationFailed  int                 `json:"verification_failed,omitempty"`
-	VerificationOutcome Outcome             `json:"verification_outcome,omitempty"`
-	Constraints         []string            `json:"constraints,omitempty"`
-	ChangedFiles        []string            `json:"changed_files,omitempty"`
-	UnresolvedFailures  []taskFailureState  `json:"unresolved_failures,omitempty"`
-	Approvals           []taskApprovalState `json:"approvals,omitempty"`
-	Artifacts           []taskArtifactState `json:"artifacts,omitempty"`
-}
-
 type preservedEdit struct {
 	Path string `json:"path"`
 	Note string `json:"note,omitempty"`
@@ -481,39 +464,8 @@ type preservedInstruction struct {
 // may live only inside the injected summary message, which on a later compaction
 // lands in middle with no real tool calls left to extract. Fresh tool calls and
 // instruction blocks override the carried-forward copy by name/source.
-func appendPreservedState(summary string, middle []zeroruntime.Message, taskSnapshot *taskStateSnapshot) string {
+func appendPreservedState(summary string, middle []zeroruntime.Message) string {
 	priorState := parsePreservedStateBlock(latestSummaryContent(middle))
-	task := priorState.Task
-	if taskSnapshot != nil {
-		task = &preservedTaskState{
-			Objective:          capTaskObjective(taskSnapshot.Objective),
-			Constraints:        mergeBoundedComparable(priorTaskConstraints(priorState.Task), taskSnapshot.Constraints, maxTaskConstraints),
-			ChangedFiles:       mergeBoundedComparable(priorTaskChangedFiles(priorState.Task), taskSnapshot.ChangedFiles, maxTaskEvidenceEntries),
-			UnresolvedFailures: mergeTaskFailures(priorTaskFailures(priorState.Task), taskSnapshot.UnresolvedFailures, taskSnapshot.ResolvedFailureKeys),
-			Approvals:          mergeBoundedComparable(priorTaskApprovals(priorState.Task), taskSnapshot.Approvals, maxTaskEvidenceEntries),
-			Artifacts:          mergeBoundedComparable(priorTaskArtifacts(priorState.Task), taskSnapshot.Artifacts, maxTaskEvidenceEntries),
-		}
-		// Plan parity corroborates only the mutable task projection. The objective
-		// comes directly from the run prompt and is immutable, so it must survive
-		// even after compaction removes the plan tool call needed for comparison.
-		if taskSnapshot.PlanParity == taskPlanParityMatch {
-			task.Status = taskSnapshot.Status
-			task.Pending = taskSnapshot.Plan.Pending
-			task.InProgress = taskSnapshot.Plan.InProgress
-			task.Completed = taskSnapshot.Plan.Completed
-			task.Failed = taskSnapshot.Plan.Failed
-			task.VerificationPassed = taskSnapshot.Verification.Passed
-			task.VerificationFailed = taskSnapshot.Verification.Failed
-			task.VerificationOutcome = taskSnapshot.Verification.LastOutcome
-		}
-	}
-	freshConstraints := explicitConstraintsFromMessages(middle)
-	if len(freshConstraints) > 0 {
-		if task == nil {
-			task = &preservedTaskState{}
-		}
-		task.Constraints = mergeBoundedComparable(task.Constraints, freshConstraints, maxTaskConstraints)
-	}
 
 	// Plan: a fresh update_plan in middle is authoritative; otherwise carry the
 	// plan preserved by an earlier compaction.
@@ -541,114 +493,12 @@ func appendPreservedState(summary string, middle []zeroruntime.Message, taskSnap
 		projectInstructionEntries(middle),
 	)
 
-	if block := formatPreservedState(plan, task, edits, tools, skills, instructions); block != "" {
+	if block := formatPreservedState(plan, edits, tools, skills, instructions); block != "" {
 		summary += "\n\n" + block
 	}
 	return summary
 }
 
-func explicitConstraintsFromMessages(messages []zeroruntime.Message) []string {
-	var constraints []string
-	for _, message := range messages {
-		if message.Role != zeroruntime.MessageRoleUser || strings.Contains(message.Content, preservedStateLabel) {
-			continue
-		}
-		if _, body := projectInstructionBlock(message.Content); body != "" {
-			continue
-		}
-		constraints = mergeBoundedComparable(constraints, extractExplicitConstraints(message.Content), maxTaskConstraints)
-	}
-	return constraints
-}
-
-func priorTaskConstraints(task *preservedTaskState) []string {
-	if task == nil {
-		return nil
-	}
-	return task.Constraints
-}
-
-func priorTaskChangedFiles(task *preservedTaskState) []string {
-	if task == nil {
-		return nil
-	}
-	return task.ChangedFiles
-}
-
-func priorTaskFailures(task *preservedTaskState) []taskFailureState {
-	if task == nil {
-		return nil
-	}
-	return task.UnresolvedFailures
-}
-
-func priorTaskApprovals(task *preservedTaskState) []taskApprovalState {
-	if task == nil {
-		return nil
-	}
-	return task.Approvals
-}
-
-func priorTaskArtifacts(task *preservedTaskState) []taskArtifactState {
-	if task == nil {
-		return nil
-	}
-	return task.Artifacts
-}
-
-func mergeTaskFailures(older, newer []taskFailureState, resolved []string) []taskFailureState {
-	resolvedSet := make(map[string]struct{}, len(resolved))
-	for _, key := range resolved {
-		resolvedSet[key] = struct{}{}
-	}
-	merged := make([]taskFailureState, 0, len(older)+len(newer))
-	for _, failure := range append(append([]taskFailureState(nil), older...), newer...) {
-		if _, ok := resolvedSet[failure.Key]; ok {
-			continue
-		}
-		withoutSameKey := make([]taskFailureState, 0, len(merged))
-		for _, existing := range merged {
-			if existing.Key != failure.Key {
-				withoutSameKey = append(withoutSameKey, existing)
-			}
-		}
-		merged = append(withoutSameKey, failure)
-	}
-	if len(merged) > maxTaskEvidenceEntries {
-		merged = merged[len(merged)-maxTaskEvidenceEntries:]
-	}
-	return merged
-}
-
-func mergeBoundedComparable[T comparable](older, newer []T, limit int) []T {
-	merged := make([]T, 0, len(older)+len(newer))
-	seen := make(map[T]struct{}, len(older)+len(newer))
-	for _, value := range append(append([]T(nil), older...), newer...) {
-		if _, ok := seen[value]; ok {
-			continue
-		}
-		seen[value] = struct{}{}
-		merged = append(merged, value)
-	}
-	if len(merged) > limit {
-		merged = merged[len(merged)-limit:]
-	}
-	return merged
-}
-
-func capTaskObjective(objective string) string {
-	objective = strings.TrimSpace(objective)
-	if len(objective) <= maxTaskObjectiveBytes {
-		return objective
-	}
-	const suffix = "…"
-	limit := maxTaskObjectiveBytes - len(suffix)
-	for limit > 0 && !utf8.RuneStart(objective[limit]) {
-		limit--
-	}
-	return strings.TrimSpace(objective[:limit]) + suffix
-}
-
 // mergeRecentEdits overlays fresh edits onto edits preserved by an earlier
 // compaction. Unlike mergeSkillEntries (which keeps refreshed entries in their
 // original slot), a path touched again by a fresh edit MOVES to the newest
@@ -705,11 +555,11 @@ func preservedEditsToEntries(edits []preservedEdit) []skillEntry {
 
 // formatPreservedState renders state as the labelled, single-line
 // JSON block. Returns "" when there is nothing to preserve.
-func formatPreservedState(plan string, task *preservedTaskState, edits, tools, skills, instructions []skillEntry) string {
-	if plan == "" && task == nil && len(edits) == 0 && len(tools) == 0 && len(skills) == 0 && len(instructions) == 0 {
+func formatPreservedState(plan string, edits, tools, skills, instructions []skillEntry) string {
+	if plan == "" && len(edits) == 0 && len(tools) == 0 && len(skills) == 0 && len(instructions) == 0 {
 		return ""
 	}
-	state := preservedState{Plan: plan, Task: task}
+	state := preservedState{Plan: plan}
 	for _, e := range edits {
 		state.RecentEdits = append(state.RecentEdits, preservedEdit{Path: e.name, Note: e.body})
 	}
@@ -729,6 +579,15 @@ func formatPreservedState(plan string, task *preservedTaskState, edits, tools, s
 	return preservedStateLabel + "\n" + string(encoded)
 }
 
+// parsePreservedState recovers the plan + skills from a prior summary's preserved
+// block. JSON escaping makes this lossless even when a skill body contains
+// markdown headings, code fences, or quotes. Returns ("", nil) when absent or
+// malformed.
+func parsePreservedState(summaryContent string) (string, []skillEntry) {
+	state := parsePreservedStateBlock(summaryContent)
+	return state.Plan, preservedSkillsToEntries(state.Skills)
+}
+
 func parsePreservedStateBlock(summaryContent string) preservedState {
 	idx := strings.LastIndex(summaryContent, preservedStateLabel)
 	if idx < 0 {

internal/agent/compaction_preserve_test.go

@@ -1,12 +1,10 @@
 package agent
 
 import (
-	"encoding/json"
 	"strings"
 	"testing"
 	"unicode/utf8"
 
-	"github.com/Gitlawb/zero/internal/tools"
 	"github.com/Gitlawb/zero/internal/zeroruntime"
 )
 
@@ -61,169 +59,6 @@ func TestCompactPreservesActivePlan(t *testing.T) {
 	}
 }
 
-func TestCompactPreservesBoundedTaskContext(t *testing.T) {
-	objective := strings.Repeat("世", maxTaskObjectiveBytes)
-	task := newTaskState(objective, nil)
-	task.observe(taskStateEvent{kind: taskStateEventPlan, arguments: `{"plan":[{"content":"write code","status":"in_progress"},{"content":"add tests","status":"pending"}]}`})
-	task.observe(taskStateEvent{kind: taskStateEventToolResult, arguments: `{"patch":"*** Update File: internal/db.go"}`, toolResult: ToolResult{
-		Name: "apply_patch", Status: tools.StatusOK, Output: "Done!", ChangedFiles: []string{"internal/db.go"},
-	}})
-	messages := stateConversation()
-	compacted, err := Compact(messages, CompactionOptions{
-		PreserveLast: 2,
-		Summarize:    func([]zeroruntime.Message) (string, error) { return "SUMMARY", nil },
-		taskState:    task.snapshotForCompaction(messages),
-	})
-	if err != nil {
-		t.Fatalf("Compact: %v", err)
-	}
-	state := parsePreservedStateBlock(compacted[1].Content)
-	if state.Task == nil || state.Task.Status != taskStatusActive || state.Task.InProgress != 1 || state.Task.Pending != 1 {
-		t.Fatalf("unexpected compact task state: %#v", state.Task)
-	}
-	if len(state.Task.Objective) > maxTaskObjectiveBytes || !utf8.ValidString(state.Task.Objective) {
-		t.Fatalf("objective was not safely bounded: %d bytes %q", len(state.Task.Objective), state.Task.Objective)
-	}
-	if len(state.Task.ChangedFiles) != 1 || state.Task.ChangedFiles[0] != "internal/db.go" {
-		t.Fatalf("changed files were not preserved: %#v", state.Task.ChangedFiles)
-	}
-}
-
-func TestCompactPreservesRuntimeEvidenceAcrossRepeatedCompaction(t *testing.T) {
-	task := newTaskState("Please keep the change focused.", nil)
-	task.observe(taskStateEvent{kind: taskStateEventToolResult, arguments: `{"cmd":"go test ./..."}`, toolResult: ToolResult{
-		Name: "exec_command", Status: tools.StatusError, Output: "Error: tests failed",
-		Meta: map[string]string{"spill_path": ".zero/artifacts/tests.txt"},
-	}})
-	task.observe(taskStateEvent{kind: taskStateEventPermission, permission: PermissionEvent{
-		ToolName: "exec_command", DecisionAction: PermissionDecisionAllowForSession, Scope: "/tmp",
-	}})
-	messages := stateConversation()
-
-	first, err := Compact(messages, CompactionOptions{
-		PreserveLast: 2,
-		Summarize:    func([]zeroruntime.Message) (string, error) { return "FIRST", nil },
-		taskState:    task.snapshotForCompaction(messages),
-	})
-	if err != nil {
-		t.Fatal(err)
-	}
-	task.observe(taskStateEvent{kind: taskStateEventToolResult, arguments: `{"cmd":"go test ./..."}`, toolResult: ToolResult{
-		Name: "exec_command", Status: tools.StatusError, Output: "Error: tests still fail with newer evidence",
-		Meta: map[string]string{"spill_path": ".zero/artifacts/tests.txt"},
-	}})
-	secondInput := append(append([]zeroruntime.Message{}, first...),
-		zeroruntime.Message{Role: zeroruntime.MessageRoleUser, Content: "Never add background memory models."},
-		zeroruntime.Message{Role: zeroruntime.MessageRoleAssistant, Content: "understood"},
-		zeroruntime.Message{Role: zeroruntime.MessageRoleUser, Content: "continue"},
-		zeroruntime.Message{Role: zeroruntime.MessageRoleAssistant, Content: "working"},
-	)
-	second, err := Compact(secondInput, CompactionOptions{
-		PreserveLast: 2,
-		Summarize:    func([]zeroruntime.Message) (string, error) { return "SECOND", nil },
-		taskState:    task.snapshotForCompaction(secondInput),
-	})
-	if err != nil {
-		t.Fatal(err)
-	}
-	state := parsePreservedStateBlock(second[1].Content)
-	if state.Task == nil || len(state.Task.UnresolvedFailures) != 1 || len(state.Task.Approvals) != 1 || len(state.Task.Artifacts) != 1 {
-		t.Fatalf("runtime evidence did not survive repeated compaction: %#v", state.Task)
-	}
-	if got := state.Task.UnresolvedFailures[0].Summary; got != "Error: tests still fail with newer evidence" {
-		t.Fatalf("newer failure evidence did not replace the older summary: %#v", state.Task.UnresolvedFailures)
-	}
-	for _, want := range []string{"Please keep the change focused.", "Never add background memory models."} {
-		if !containsString(state.Task.Constraints, want) {
-			t.Fatalf("constraint %q missing after repeated compaction: %#v", want, state.Task.Constraints)
-		}
-	}
-}
-
-func containsString(values []string, want string) bool {
-	for _, value := range values {
-		if value == want {
-			return true
-		}
-	}
-	return false
-}
-
-func TestCompactPreservesObjectiveAfterPlanParityMismatch(t *testing.T) {
-	prior := preservedState{Task: &preservedTaskState{Objective: "stale objective", Status: taskStatusActive, Pending: 1}}
-	encoded, err := json.Marshal(prior)
-	if err != nil {
-		t.Fatal(err)
-	}
-	messages := []zeroruntime.Message{
-		{Role: zeroruntime.MessageRoleSystem, Content: "system"},
-		{Role: zeroruntime.MessageRoleUser, Content: summaryLabel + "\nold\n\n" + preservedStateLabel + "\n" + string(encoded)},
-		{Role: zeroruntime.MessageRoleAssistant, Content: "continuing"},
-		{Role: zeroruntime.MessageRoleUser, Content: "more"},
-		{Role: zeroruntime.MessageRoleAssistant, Content: "working"},
-		{Role: zeroruntime.MessageRoleUser, Content: "again"},
-		{Role: zeroruntime.MessageRoleAssistant, Content: "done"},
-	}
-	compacted, err := Compact(messages, CompactionOptions{
-		PreserveLast: 2,
-		Summarize:    func([]zeroruntime.Message) (string, error) { return "SUMMARY", nil },
-		taskState: &taskStateSnapshot{
-			Objective:  "current objective",
-			PlanParity: taskPlanParityMismatch,
-		},
-	})
-	if err != nil {
-		t.Fatalf("Compact: %v", err)
-	}
-	state := parsePreservedStateBlock(compacted[1].Content)
-	if state.Task == nil || state.Task.Objective != "current objective" {
-		t.Fatalf("immutable objective was lost on plan mismatch: %#v", state.Task)
-	}
-	if state.Task.Status != "" || state.Task.Pending != 0 {
-		t.Fatalf("uncorroborated mutable task fields survived plan mismatch: %#v", state.Task)
-	}
-}
-
-func TestTaskObjectiveSurvivesRepeatedCompactionWithoutPlanRefresh(t *testing.T) {
-	task := newTaskState("keep this objective", nil)
-	task.observe(taskStateEvent{kind: taskStateEventPlan, arguments: `{"plan":[{"content":"write code","status":"in_progress"},{"content":"add tests","status":"pending"}]}`})
-	messages := stateConversation()
-
-	first, err := Compact(messages, CompactionOptions{
-		PreserveLast: 2,
-		Summarize:    func([]zeroruntime.Message) (string, error) { return "FIRST", nil },
-		taskState:    task.snapshotForCompaction(messages),
-	})
-	if err != nil {
-		t.Fatalf("first Compact: %v", err)
-	}
-	secondInput := append(append([]zeroruntime.Message{}, first...),
-		zeroruntime.Message{Role: zeroruntime.MessageRoleUser, Content: "more"},
-		zeroruntime.Message{Role: zeroruntime.MessageRoleAssistant, Content: "working"},
-		zeroruntime.Message{Role: zeroruntime.MessageRoleUser, Content: "again"},
-		zeroruntime.Message{Role: zeroruntime.MessageRoleAssistant, Content: "done"},
-	)
-	snapshot := task.snapshotForCompaction(secondInput)
-	if snapshot.PlanParity != taskPlanParityMismatch {
-		t.Fatalf("plan call should be absent after first compaction, parity=%q", snapshot.PlanParity)
-	}
-	second, err := Compact(secondInput, CompactionOptions{
-		PreserveLast: 2,
-		Summarize:    func([]zeroruntime.Message) (string, error) { return "SECOND", nil },
-		taskState:    snapshot,
-	})
-	if err != nil {
-		t.Fatalf("second Compact: %v", err)
-	}
-	state := parsePreservedStateBlock(second[1].Content)
-	if state.Task == nil || state.Task.Objective != "keep this objective" {
-		t.Fatalf("objective lost after second compaction: %#v", state.Task)
-	}
-	if state.Task.Status != "" || state.Task.InProgress != 0 {
-		t.Fatalf("uncorroborated mutable fields survived second compaction: %#v", state.Task)
-	}
-}
-
 func TestCompactPreservesLoadedSkills(t *testing.T) {
 	summary := compactStateConversation(t, stateConversation())
 	if !strings.Contains(summary, preservedStateLabel) {
@@ -448,10 +283,12 @@ func TestExtractLatestPlanReturnsMostRecent(t *testing.T) {
 	}
 }
 
-func TestFormatPlanArgumentsRejectsUnsupportedStepAlias(t *testing.T) {
+func TestFormatPlanArgumentsAcceptsStepAlias(t *testing.T) {
 	got := formatPlanArguments(`{"plan":[{"step":"write failing test","status":"in_progress"},{"content":"keep existing shape","status":"pending"}]}`)
-	if got != "" {
-		t.Fatalf("unsupported alias should reject the whole plan like update_plan, got %q", got)
+	for _, want := range []string{"- [in_progress] write failing test", "- [pending] keep existing shape"} {
+		if !strings.Contains(got, want) {
+			t.Fatalf("expected %q in formatted plan, got %q", want, got)
+		}
 	}
 }
 

internal/agent/compaction_projection.go

@@ -1,315 +0,0 @@
-package agent
-
-import (
-	"encoding/json"
-	"fmt"
-	"regexp"
-	"strings"
-
-	"github.com/Gitlawb/zero/internal/zeroruntime"
-)
-
-const (
-	compactionUserWordBudget       = 256
-	compactionAssistantWordBudget  = 200
-	compactionToolCallsPerTurn     = 8
-	compactionToolArgumentBytes    = 120
-	compactionErrorBytes           = 150
-	compactionResultBytes          = 1200
-	compactionPreviousSummaryBytes = 16 * 1024
-	compactionBriefMaxBytes        = 24 * 1024
-)
-
-var compactionWordPattern = regexp.MustCompile(`[\p{L}\p{N}]+`)
-
-type compactionProjection struct {
-	messages  []zeroruntime.Message
-	truncated bool
-}
-
-// projectCompactionInput removes reconstructible tool output before the model
-// summary call. It retains a bounded transcript of user intent, assistant
-// decisions, tool actions, and concise errors. Exact durable state is appended
-// separately from the original messages after the summary returns.
-func projectCompactionInput(messages []zeroruntime.Message) compactionProjection {
-	sections := make([]string, 0, len(messages))
-	previousSummary := ""
-	truncated := false
-	for index, message := range messages {
-		switch message.Role {
-		case zeroruntime.MessageRoleUser:
-			content := message.Content
-			if marker := strings.Index(content, preservedStateLabel); marker >= 0 {
-				content = content[:marker]
-			}
-			trimmed := strings.TrimSpace(content)
-			if strings.HasPrefix(trimmed, summaryLabel) {
-				fullSummary := strings.TrimSpace(strings.TrimPrefix(trimmed, summaryLabel))
-				previousSummary = clipHeadTailBytes(fullSummary, compactionPreviousSummaryBytes)
-				truncated = truncated || len(fullSummary) > compactionPreviousSummaryBytes
-				continue
-			}
-			if content = clipInformativeWords(content, compactionUserWordBudget); content != "" {
-				sections = append(sections, fmt.Sprintf("[user #%d]\n%s", index, content))
-			}
-		case zeroruntime.MessageRoleAssistant:
-			lines := make([]string, 0, 1+len(message.ToolCalls))
-			if content := clipInformativeWords(message.Content, compactionAssistantWordBudget); content != "" {
-				lines = append(lines, content)
-			}
-			calls := message.ToolCalls
-			omitted := 0
-			if len(calls) > compactionToolCallsPerTurn {
-				omitted = len(calls) - compactionToolCallsPerTurn
-				calls = calls[omitted:]
-			}
-			if omitted > 0 {
-				lines = append(lines, fmt.Sprintf("* (%d earlier tool calls omitted)", omitted))
-			}
-			for _, call := range calls {
-				lines = append(lines, compactionToolCallLine(call))
-			}
-			if len(lines) > 0 {
-				sections = append(sections, fmt.Sprintf("[assistant #%d]\n%s", index, strings.Join(lines, "\n")))
-			}
-		case zeroruntime.MessageRoleTool:
-			// New tool-result messages carry the execution status directly, as the
-			// source ToolResult does. Keep the text check for older session history
-			// created before Message exposed IsError.
-			name := toolNameForResult(messages, index)
-			mustPreserve := name == "ask_user" || len(message.ChangedFiles) > 0
-			if !mustPreserve && !message.IsError && !isLikelyToolError(message.Content) {
-				continue
-			}
-			if name == "" {
-				name = "tool"
-			}
-			switch {
-			case message.IsError || isLikelyToolError(message.Content):
-				sections = append(sections, fmt.Sprintf("[tool_error #%d] %s\n%s", index, name, clipBytes(firstLine(message.Content), compactionErrorBytes)))
-			case name == "ask_user":
-				sections = append(sections, fmt.Sprintf("[user_answer #%d] %s\n%s", index, name, clipBytes(message.Content, compactionResultBytes)))
-			default:
-				sections = append(sections, fmt.Sprintf("[tool_result #%d] %s changed %s\n%s", index, name,
-					clipBytes(strings.Join(message.ChangedFiles, ", "), compactionToolArgumentBytes), clipBytes(firstLine(message.Content), compactionErrorBytes)))
-			}
-		}
-	}
-	if len(sections) == 0 && previousSummary == "" {
-		return compactionProjection{}
-	}
-	brief, briefTruncated := capCompactionBrief(strings.Join(sections, "\n\n"))
-	truncated = truncated || briefTruncated
-	if previousSummary != "" {
-		brief = "[previous summary]\n" + previousSummary + "\n\n" + brief
-	}
-	return compactionProjection{
-		messages:  []zeroruntime.Message{{Role: zeroruntime.MessageRoleUser, Content: strings.TrimSpace(brief)}},
-		truncated: truncated,
-	}
-}
-
-func capCompactionBrief(brief string) (string, bool) {
-	if len(brief) <= compactionBriefMaxBytes {
-		return brief, false
-	}
-	marker := "\n\n...[middle transcript omitted to fit compaction budget]...\n\n"
-	available := compactionBriefMaxBytes - len(marker)
-	// Keep both chronological edges: the oldest material has not necessarily
-	// been summarized before, while the newest material is most actionable.
-	headBytes := available * 2 / 5
-	tailBytes := available - headBytes
-	head := clipPrefixAtBoundary(brief, headBytes)
-	tail := clipSuffixAtBoundary(brief, tailBytes)
-	return strings.TrimSpace(head) + marker + strings.TrimSpace(tail), true
-}
-
-func compactionToolCallLine(call zeroruntime.ToolCall) string {
-	detail := ""
-	var arguments map[string]any
-	if json.Unmarshal([]byte(strings.TrimSpace(call.Arguments)), &arguments) == nil {
-		for _, key := range []string{"file_path", "path", "url", "query", "pattern", "prompt", "description", "question"} {
-			if value, ok := arguments[key].(string); ok && strings.TrimSpace(value) != "" {
-				detail = value
-				break
-			}
-		}
-		if detail == "" && call.Name == "ask_user" {
-			detail = askUserQuestionsDetail(arguments)
-		}
-		if detail == "" && call.Name == "apply_patch" {
-			if patch, ok := arguments["patch"].(string); ok {
-				detail = strings.Join(patchPaths(patch), ", ")
-			}
-		}
-	}
-	if detail == "" {
-		detail = commandFromArguments(call.Arguments)
-	}
-	if detail == "" {
-		return "* " + call.Name
-	}
-	return fmt.Sprintf("* %s %q", call.Name, clipBytes(detail, compactionToolArgumentBytes))
-}
-
-func isLikelyToolError(content string) bool {
-	prefix := strings.TrimSpace(content)
-	if len(prefix) > compactionErrorBytes {
-		prefix = prefix[:compactionErrorBytes]
-	}
-	lower := strings.ToLower(prefix)
-	for _, prefix := range []string{"error:", "tool error:", "failed:", "permission denied", "command failed"} {
-		if strings.HasPrefix(lower, prefix) {
-			return true
-		}
-	}
-	return false
-}
-
-func clipInformativeWords(text string, limit int) string {
-	flat := strings.Join(strings.Fields(text), " ")
-	if flat == "" || limit <= 0 {
-		return ""
-	}
-	count := 0
-	for _, location := range compactionWordPattern.FindAllStringIndex(flat, -1) {
-		word := strings.ToLower(flat[location[0]:location[1]])
-		if compactionStopWords[word] {
-			continue
-		}
-		count++
-		if count > limit {
-			end := location[0]
-			return strings.TrimSpace(flat[:end]) + "...(truncated)"
-		}
-	}
-	return flat
-}
-
-func clipBytes(text string, limit int) string {
-	text = strings.TrimSpace(text)
-	if limit <= 0 {
-		return ""
-	}
-	if len(text) <= limit {
-		return text
-	}
-	if limit <= 3 {
-		end := limit
-		for end > 0 && end < len(text) && text[end]&0xc0 == 0x80 {
-			end--
-		}
-		return text[:end]
-	}
-	end := limit - 3
-	for end > 0 && end < len(text) && text[end]&0xc0 == 0x80 {
-		end--
-	}
-	return strings.TrimSpace(text[:end]) + "..."
-}
-
-func clipHeadTailBytes(text string, limit int) string {
-	text = strings.TrimSpace(text)
-	if len(text) <= limit {
-		return text
-	}
-	if limit <= 0 {
-		return ""
-	}
-	marker := "\n...[middle omitted]...\n"
-	if limit <= len(marker) {
-		return clipBytes(text, limit)
-	}
-	available := limit - len(marker)
-	head := clipPrefixAtBoundary(text, available/2)
-	tail := clipSuffixAtBoundary(text, available-available/2)
-	return strings.TrimSpace(head) + marker + strings.TrimSpace(tail)
-}
-
-func clipPrefixAtBoundary(text string, limit int) string {
-	if limit <= 0 {
-		return ""
-	}
-	if len(text) <= limit {
-		return text
-	}
-	end := limit
-	for end > 0 && text[end]&0xc0 == 0x80 {
-		end--
-	}
-	if boundary := strings.LastIndex(text[:end], "\n\n["); boundary > 0 {
-		end = boundary
-	}
-	return text[:end]
-}
-
-func clipSuffixAtBoundary(text string, limit int) string {
-	if limit <= 0 {
-		return ""
-	}
-	if len(text) <= limit {
-		return text
-	}
-	start := len(text) - limit
-	for start < len(text) && text[start]&0xc0 == 0x80 {
-		start++
-	}
-	if boundary := strings.Index(text[start:], "\n\n["); boundary >= 0 {
-		start += boundary + 2
-	}
-	return text[start:]
-}
-
-func askUserQuestionsDetail(arguments map[string]any) string {
-	raw, ok := arguments["questions"].([]any)
-	if !ok {
-		return ""
-	}
-	questions := make([]string, 0, len(raw))
-	for _, item := range raw {
-		object, ok := item.(map[string]any)
-		if !ok {
-			continue
-		}
-		if question, ok := object["question"].(string); ok && strings.TrimSpace(question) != "" {
-			questions = append(questions, strings.TrimSpace(question))
-		}
-	}
-	return strings.Join(questions, " | ")
-}
-
-func patchPaths(patch string) []string {
-	var paths []string
-	seen := map[string]bool{}
-	for _, line := range strings.Split(patch, "\n") {
-		for _, prefix := range []string{"*** Add File: ", "*** Update File: ", "*** Delete File: ", "*** Move to: "} {
-			if !strings.HasPrefix(line, prefix) {
-				continue
-			}
-			path := strings.TrimSpace(strings.TrimPrefix(line, prefix))
-			if path != "" && !seen[path] {
-				seen[path] = true
-				paths = append(paths, path)
-			}
-		}
-	}
-	return paths
-}
-
-var compactionStopWords = map[string]bool{
-	"a": true, "an": true, "the": true, "is": true, "are": true, "was": true, "were": true,
-	"be": true, "been": true, "being": true, "have": true, "has": true, "had": true,
-	"do": true, "does": true, "did": true, "will": true, "would": true, "could": true,
-	"should": true, "may": true, "might": true, "shall": true, "can": true, "need": true, "must": true,
-	"to": true, "of": true, "in": true, "for": true, "on": true, "with": true, "at": true,
-	"by": true, "from": true, "as": true, "into": true, "through": true, "during": true,
-	"before": true, "after": true, "above": true, "below": true, "between": true, "under": true, "over": true,
-	"and": true, "but": true, "or": true, "nor": true, "not": true, "so": true, "yet": true,
-	"both": true, "either": true, "neither": true, "each": true, "every": true, "all": true,
-	"any": true, "few": true, "more": true, "most": true, "other": true, "some": true, "such": true, "no": true,
-	"that": true, "this": true, "these": true, "those": true, "it": true, "its": true,
-	"i": true, "me": true, "my": true, "we": true, "our": true, "you": true, "your": true,
-	"he": true, "him": true, "his": true, "she": true, "her": true, "they": true, "them": true, "their": true,
-	"who": true, "which": true, "what": true, "if": true, "then": true, "than": true,
-	"when": true, "where": true, "how": true, "just": true, "also": true,
-}

internal/agent/compaction_projection_test.go

@@ -1,130 +0,0 @@
-package agent
-
-import (
-	"strings"
-	"testing"
-	"unicode/utf8"
-
-	"github.com/Gitlawb/zero/internal/zeroruntime"
-)
-
-func TestClipBytesHandlesTinyLimitsOnRuneBoundaries(t *testing.T) {
-	for _, limit := range []int{0, 1, 2, 3} {
-		if clipped := clipBytes("世界", limit); len(clipped) > limit || !utf8.ValidString(clipped) {
-			t.Fatalf("clipBytes limit %d returned invalid result %q", limit, clipped)
-		}
-	}
-}
-
-func TestBoundaryClippersHandleNonPositiveLimits(t *testing.T) {
-	for _, clip := range []func(string, int) string{clipPrefixAtBoundary, clipSuffixAtBoundary} {
-		for _, limit := range []int{-1, 0} {
-			if got := clip("世界", limit); got != "" {
-				t.Fatalf("clipper limit %d = %q, want empty", limit, got)
-			}
-		}
-	}
-}
-
-func TestProjectCompactionInputDropsRecoverableToolBodies(t *testing.T) {
-	largeBody := strings.Repeat("recoverable source line\n", 1000)
-	messages := []zeroruntime.Message{
-		{Role: zeroruntime.MessageRoleUser, Content: "Inspect the parser and preserve its current behavior."},
-		{Role: zeroruntime.MessageRoleAssistant, Content: "I will inspect before editing.", ToolCalls: []zeroruntime.ToolCall{{ID: "read", Name: "read_file", Arguments: `{"path":"internal/parser.go"}`}}},
-		{Role: zeroruntime.MessageRoleTool, ToolCallID: "read", Content: largeBody},
-		{Role: zeroruntime.MessageRoleAssistant, ToolCalls: []zeroruntime.ToolCall{{ID: "test", Name: "exec_command", Arguments: `{"cmd":"go test ./internal/parser"}`}}},
-		{Role: zeroruntime.MessageRoleTool, ToolCallID: "test", Content: "Error: parser regression\nlong diagnostic details"},
-	}
-
-	projection := projectCompactionInput(messages)
-	projected := projection.messages
-	if len(projected) != 1 {
-		t.Fatalf("projection = %#v, want one compact message", projected)
-	}
-	content := projected[0].Content
-	for _, want := range []string{"Inspect the parser", `read_file "internal/parser.go"`, `exec_command "go test ./internal/parser"`, "Error: parser regression"} {
-		if !strings.Contains(content, want) {
-			t.Fatalf("projection missing %q: %q", want, content)
-		}
-	}
-	if strings.Contains(content, "recoverable source line") || strings.Contains(content, "long diagnostic details") {
-		t.Fatalf("projection retained reconstructible or overlong tool output: %q", content)
-	}
-	if estimateTokens(projected)*20 >= estimateTokens(messages) {
-		t.Fatalf("projection did not materially reduce tokens: before=%d after=%d", estimateTokens(messages), estimateTokens(projected))
-	}
-}
-
-func TestProjectCompactionInputRetainsStructurallyMarkedToolErrors(t *testing.T) {
-	messages := []zeroruntime.Message{
-		{Role: zeroruntime.MessageRoleAssistant, ToolCalls: []zeroruntime.ToolCall{{ID: "git", Name: "exec_command", Arguments: `{"cmd":"git status"}`}}},
-		{Role: zeroruntime.MessageRoleTool, ToolCallID: "git", Content: "fatal: not a git repository", IsError: true},
-	}
-
-	projected := projectCompactionInput(messages).messages
-	if len(projected) != 1 || !strings.Contains(projected[0].Content, "fatal: not a git repository") {
-		t.Fatalf("structured tool error was dropped from projection: %#v", projected)
-	}
-}
-
-func TestProjectCompactionInputRetainsAskUserExchangeAndSemanticArguments(t *testing.T) {
-	messages := []zeroruntime.Message{
-		{Role: zeroruntime.MessageRoleAssistant, ToolCalls: []zeroruntime.ToolCall{{ID: "ask", Name: "ask_user", Arguments: `{"questions":[{"question":"Which database should remain?"}]}`}}},
-		{Role: zeroruntime.MessageRoleTool, ToolCallID: "ask", Content: "Which database should remain?: Postgres only; do not touch MySQL."},
-		{Role: zeroruntime.MessageRoleAssistant, ToolCalls: []zeroruntime.ToolCall{
-			{ID: "patch", Name: "apply_patch", Arguments: `{"patch":"*** Begin Patch\n*** Update File: internal/db.go\n*** End Patch"}`},
-			{ID: "fetch", Name: "web_fetch", Arguments: `{"url":"https://example.com/spec"}`},
-			{ID: "task", Name: "task", Arguments: `{"prompt":"Inspect the Windows implementation"}`},
-		}},
-		{Role: zeroruntime.MessageRoleTool, ToolCallID: "patch", Content: "Done!", ChangedFiles: []string{"internal/db.go"}},
-	}
-
-	content := projectCompactionInput(messages).messages[0].Content
-	for _, want := range []string{"Which database should remain?", "Postgres only; do not touch MySQL", "internal/db.go", "https://example.com/spec", "Inspect the Windows implementation"} {
-		if !strings.Contains(content, want) {
-			t.Fatalf("projection missing %q: %q", want, content)
-		}
-	}
-}
-
-func TestProjectCompactionInputCapsToolCallsPerTurn(t *testing.T) {
-	calls := make([]zeroruntime.ToolCall, 12)
-	for index := range calls {
-		calls[index] = zeroruntime.ToolCall{Name: "read_file", Arguments: `{"path":"file.go"}`}
-	}
-	projected := projectCompactionInput([]zeroruntime.Message{{Role: zeroruntime.MessageRoleAssistant, ToolCalls: calls}}).messages
-	content := projected[0].Content
-	if strings.Count(content, "* read_file") != compactionToolCallsPerTurn || !strings.Contains(content, "4 earlier tool calls omitted") {
-		t.Fatalf("tool-call tail was not bounded: %q", content)
-	}
-}
-
-func TestProjectCompactionInputCarriesPreviousSummaryOutsideBriefCap(t *testing.T) {
-	previous := "keep the earlier architecture decision\n" + strings.Repeat("prior detail ", 2000) + "\nkeep the newest prior decision"
-	messages := []zeroruntime.Message{{Role: zeroruntime.MessageRoleUser, Content: summaryLabel + "\n" + previous + "\n\n" + preservedStateLabel + "\n{}"}}
-	for index := range 500 {
-		messages = append(messages, zeroruntime.Message{Role: zeroruntime.MessageRoleAssistant, Content: strings.Repeat("later transcript detail ", 20)})
-		messages = append(messages, zeroruntime.Message{Role: zeroruntime.MessageRoleUser, Content: "later request " + strings.Repeat("context ", 20) + string(rune('a'+index%26))})
-	}
-
-	projection := projectCompactionInput(messages)
-	content := projection.messages[0].Content
-	if !strings.Contains(content, "[previous summary]\nkeep the earlier architecture decision") || !strings.Contains(content, "keep the newest prior decision") {
-		t.Fatalf("previous summary was lost when the brief tail was capped: %q", content)
-	}
-	if !strings.Contains(content, "middle transcript omitted to fit compaction budget") {
-		t.Fatalf("oversized brief was not head/tail capped: %q", content)
-	}
-	if !projection.truncated {
-		t.Fatal("projection did not structurally report truncation")
-	}
-}
-
-func TestProjectCompactionInputDoesNotInferTruncationFromUserText(t *testing.T) {
-	projection := projectCompactionInput([]zeroruntime.Message{{
-		Role: zeroruntime.MessageRoleUser, Content: "Keep this literal marker: [middle omitted]",
-	}})
-	if projection.truncated {
-		t.Fatal("literal user text produced a false truncation signal")
-	}
-}

internal/agent/compaction_recent_edits_test.go

@@ -20,15 +20,11 @@ func TestRecentEditsExtractsPathsAndNotes(t *testing.T) {
 			{ID: "e2", Name: "edit_file", Arguments: `{"path":"internal/bar.go","old_string":"a","new_string":"b"}`},
 		}},
 		{Role: zeroruntime.MessageRoleTool, ToolCallID: "e2", Content: "Applied edit to internal/bar.go"},
-		{Role: zeroruntime.MessageRoleAssistant, ToolCalls: []zeroruntime.ToolCall{
-			{ID: "e3", Name: "apply_patch", Arguments: `{"patch":"*** Update File: internal/baz.go"}`},
-		}},
-		{Role: zeroruntime.MessageRoleTool, ToolCallID: "e3", Content: "Done!", ChangedFiles: []string{"internal/baz.go"}},
 	}
 
 	edits := recentEdits(messages)
-	if len(edits) != 3 {
-		t.Fatalf("expected 3 edited files, got %d: %#v", len(edits), edits)
+	if len(edits) != 2 {
+		t.Fatalf("expected 2 edited files, got %d: %#v", len(edits), edits)
 	}
 	if edits[0].name != "internal/foo.go" || !strings.Contains(edits[0].body, "12 lines") {
 		t.Fatalf("first edit = %#v, want foo.go with its note", edits[0])
@@ -36,9 +32,6 @@ func TestRecentEditsExtractsPathsAndNotes(t *testing.T) {
 	if edits[1].name != "internal/bar.go" || !strings.Contains(edits[1].body, "Applied edit") {
 		t.Fatalf("second edit = %#v, want bar.go with its note", edits[1])
 	}
-	if edits[2].name != "internal/baz.go" || edits[2].body != "Done!" {
-		t.Fatalf("third edit = %#v, want apply_patch changed file with its note", edits[2])
-	}
 }
 
 // After compaction elides the editing turns, the preserved-state block still

internal/agent/compaction_test.go

@@ -7,7 +7,6 @@ import (
 	"testing"
 
 	"github.com/Gitlawb/zero/internal/tools"
-	"github.com/Gitlawb/zero/internal/trace"
 	"github.com/Gitlawb/zero/internal/zeroruntime"
 )
 
@@ -59,15 +58,12 @@ func TestCompactKeepsSystemAndPreservedSuffix(t *testing.T) {
 	if last.Content != "most recent question" {
 		t.Fatalf("expected most recent message preserved, got %q", last.Content)
 	}
-	// The summarizer receives a compact semantic projection of the middle rather
-	// than reconstructible raw tool output.
-	if len(captured) != 1 || captured[0].Role != zeroruntime.MessageRoleUser {
-		t.Fatalf("expected one projected summary input, got %#v", captured)
+	// The summarized middle excludes system and preserved suffix.
+	if len(captured) != 3 {
+		t.Fatalf("expected 3 summarized messages, got %d: %#v", len(captured), captured)
 	}
-	for _, want := range []string{"[user #0]", "first question", "[assistant #1]", "first answer", "second question"} {
-		if !strings.Contains(captured[0].Content, want) {
-			t.Fatalf("projected summary input missing %q: %q", want, captured[0].Content)
-		}
+	if captured[0].Content != "first question" {
+		t.Fatalf("expected oldest non-system message first, got %#v", captured[0])
 	}
 	// Compaction must shrink the conversation.
 	if estimateTokens(result) >= estimateTokens(messages) {
@@ -340,7 +336,7 @@ func TestRunProactiveCompactionTriggers(t *testing.T) {
 	}
 
 	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedReadFileTool(t.TempDir(), nil))
+	registry.Register(tools.NewReadFileTool(t.TempDir()))
 
 	result, err := Run(context.Background(), strings.Repeat("y", 8000), provider, Options{
 		Registry:               registry,
@@ -368,7 +364,7 @@ func TestRunNoCompactionWhenContextWindowZero(t *testing.T) {
 		},
 	}
 	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedReadFileTool(t.TempDir(), nil))
+	registry.Register(tools.NewReadFileTool(t.TempDir()))
 
 	_, err := Run(context.Background(), strings.Repeat("y", 8000), provider, Options{
 		Registry:       registry,
@@ -383,43 +379,6 @@ func TestRunNoCompactionWhenContextWindowZero(t *testing.T) {
 	}
 }
 
-func TestMaybeCompactStopsAfterSupersededReadsReachThreshold(t *testing.T) {
-	body := strings.Repeat("same source line\n", 1000)
-	messages := []zeroruntime.Message{
-		{Role: zeroruntime.MessageRoleSystem, Content: "system"},
-		toolCallWithArgs("old", "read_file", `{"path":"a.go"}`),
-		{Role: zeroruntime.MessageRoleTool, ToolCallID: "old", Content: body},
-		toolCallWithArgs("new", "read_file", `{"path":"a.go"}`),
-		{Role: zeroruntime.MessageRoleTool, ToolCallID: "new", Content: body},
-	}
-	pruned, reclaimed := pruneSupersededReadResults(messages)
-	if reclaimed <= 0 {
-		t.Fatal("test setup did not reclaim superseded read output")
-	}
-	prunedSize := estimateTokens(pruned)
-	if estimateTokens(messages) <= prunedSize {
-		t.Fatal("test setup does not cross the compaction threshold")
-	}
-
-	provider := &summarizeRecordingProvider{}
-	state := &compactionState{
-		enabled:      true,
-		threshold:    prunedSize,
-		preserveLast: 2,
-	}
-	got := state.maybeCompact(context.Background(), provider, messages, nil)
-
-	if provider.summarizeCalls != 0 {
-		t.Fatalf("cheap pruning should avoid the summarizer, got %d calls", provider.summarizeCalls)
-	}
-	if state.lowWaterMark != prunedSize {
-		t.Fatalf("lowWaterMark = %d, want %d", state.lowWaterMark, prunedSize)
-	}
-	if !strings.Contains(got[2].Content, "superseded identical") || got[4].Content != body {
-		t.Fatalf("unexpected pruned messages: %#v", got)
-	}
-}
-
 // reactiveProvider builds up history with a tool call on turn 1, then errors
 // with a context-limit message on turn 2 (once the history is large enough that
 // compaction can shrink it), then succeeds on the same-turn retry. Summary
@@ -429,7 +388,6 @@ type reactiveProvider struct {
 	summarizeCalls int
 	turnRequests   int
 	failedOnce     bool
-	connectError   bool
 	bigText        string
 	finalText      string
 }
@@ -451,9 +409,6 @@ func (provider *reactiveProvider) StreamCompletion(ctx context.Context, request
 		return streamEvents(toolTurnWithText(provider.bigText, "1", "read_file", `{"path":"x"}`)), nil
 	case provider.turnRequests == 2 && !provider.failedOnce:
 		provider.failedOnce = true
-		if provider.connectError {
-			return nil, errors.New("context_length_exceeded")
-		}
 		return streamEvents([]zeroruntime.StreamEvent{
 			{Type: zeroruntime.StreamEventError, Error: "This model's maximum context length is 1000 tokens. Please reduce the length of the messages."},
 		}), nil
@@ -462,39 +417,6 @@ func (provider *reactiveProvider) StreamCompletion(ctx context.Context, request
 	}
 }
 
-func TestRunConnectErrorCompactionRecordsReplacementPlan(t *testing.T) {
-	provider := &reactiveProvider{
-		connectError: true,
-		bigText:      strings.Repeat("b", 6000),
-		finalText:    "recovered",
-	}
-	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedReadFileTool(t.TempDir(), nil))
-	recorder := trace.NewRecorder("connect-compaction", "run-1", "test")
-	var contextPlans []ContextBreakdown
-
-	result, err := Run(context.Background(), strings.Repeat("z", 6000), provider, Options{
-		Registry:               registry,
-		PermissionMode:         PermissionModeUnsafe,
-		ContextWindow:          10_000_000,
-		CompactionPreserveLast: 2,
-		Trace:                  recorder,
-		OnContext: func(breakdown ContextBreakdown) {
-			contextPlans = append(contextPlans, breakdown)
-		},
-	})
-	if err != nil || result.FinalAnswer != "recovered" {
-		t.Fatalf("connect-error compaction failed: result=%#v err=%v", result, err)
-	}
-	if len(contextPlans) != 3 || contextPlans[2].MessageTokens >= contextPlans[1].MessageTokens {
-		t.Fatalf("replacement request context not reported: %#v", contextPlans)
-	}
-	gotTrace := recorder.Finish()
-	if len(gotTrace.PrefixHashes) != 4 || gotTrace.PrefixHashes[3].InvalidationReason != "unchanged" {
-		t.Fatalf("replacement request trace not recorded: %#v", gotTrace.PrefixHashes)
-	}
-}
-
 func TestRunReactiveCompactionRecovers(t *testing.T) {
 	provider := &reactiveProvider{
 		bigText:   strings.Repeat("b", 6000),
@@ -505,9 +427,7 @@ func TestRunReactiveCompactionRecovers(t *testing.T) {
 	// the compaction closure. read_file also returns a sizeable result that
 	// bloats the history.
 	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedReadFileTool(t.TempDir(), nil))
-	recorder := trace.NewRecorder("reactive-compaction", "run-1", "test")
-	var contextPlans []ContextBreakdown
+	registry.Register(tools.NewReadFileTool(t.TempDir()))
 
 	// ContextWindow large enough that proactive compaction never triggers, so
 	// only the reactive path can save the run.
@@ -516,10 +436,6 @@ func TestRunReactiveCompactionRecovers(t *testing.T) {
 		PermissionMode:         PermissionModeUnsafe,
 		ContextWindow:          10_000_000,
 		CompactionPreserveLast: 2,
-		Trace:                  recorder,
-		OnContext: func(breakdown ContextBreakdown) {
-			contextPlans = append(contextPlans, breakdown)
-		},
 	})
 	if err != nil {
 		t.Fatalf("expected reactive compaction to recover the run, got error: %v", err)
@@ -530,25 +446,6 @@ func TestRunReactiveCompactionRecovers(t *testing.T) {
 	if provider.summarizeCalls == 0 {
 		t.Fatal("expected reactive compaction to invoke the summarizer")
 	}
-	if len(contextPlans) != 3 {
-		t.Fatalf("main request context callbacks = %d, want initial turns plus compacted retry", len(contextPlans))
-	}
-	if contextPlans[2].MessageTokens >= contextPlans[1].MessageTokens {
-		t.Fatalf("retry context was not updated after compaction: before=%d after=%d", contextPlans[1].MessageTokens, contextPlans[2].MessageTokens)
-	}
-	gotTrace := recorder.Finish()
-	if len(gotTrace.PrefixHashes) != 4 {
-		t.Fatalf("prefix evidence count = %d, want two turns, compaction, and retry: %#v", len(gotTrace.PrefixHashes), gotTrace.PrefixHashes)
-	}
-	if gotTrace.PrefixHashes[0].InvalidationReason != "initial" ||
-		gotTrace.PrefixHashes[1].InvalidationReason != "unchanged" ||
-		gotTrace.PrefixHashes[2].InvalidationReason != "initial" ||
-		gotTrace.PrefixHashes[3].InvalidationReason != "unchanged" {
-		t.Fatalf("unexpected request-specific invalidation evidence: %#v", gotTrace.PrefixHashes)
-	}
-	if gotTrace.PrefixHashes[2].CompletePrefixHash == gotTrace.PrefixHashes[1].CompletePrefixHash {
-		t.Fatalf("compaction fingerprint reused the main request prefix: %#v", gotTrace.PrefixHashes)
-	}
 }
 
 // midStreamReactiveProvider forwards some text BEFORE surfacing a context-limit
@@ -594,7 +491,7 @@ func TestRunReactiveRetryDoesNotDoubleEmitText(t *testing.T) {
 		finalText:   "recovered",
 	}
 	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedReadFileTool(t.TempDir(), nil))
+	registry.Register(tools.NewReadFileTool(t.TempDir()))
 
 	var deltas []string
 	result, err := Run(context.Background(), strings.Repeat("z", 6000), provider, Options{
@@ -684,7 +581,7 @@ func TestCompactNeverProducesConsecutiveUserMessages(t *testing.T) {
 }
 
 func TestRecoverNoopDoesNotConsumeReactiveBudget(t *testing.T) {
-	st := newCompactionState(Options{ContextWindow: 1000, CompactionPreserveLast: 2}, nil)
+	st := newCompactionState(Options{ContextWindow: 1000, CompactionPreserveLast: 2})
 	provider := &mockProvider{turns: [][]zeroruntime.StreamEvent{{
 		{Type: zeroruntime.StreamEventText, Content: "SUMMARY"}, {Type: zeroruntime.StreamEventDone},
 	}}}
@@ -731,7 +628,7 @@ func TestRecoverNoopDoesNotConsumeReactiveBudget(t *testing.T) {
 }
 
 func TestRecoverDisabledIsNoop(t *testing.T) {
-	st := newCompactionState(Options{ContextWindow: 0}, nil)
+	st := newCompactionState(Options{ContextWindow: 0})
 	msgs := []zeroruntime.Message{{Role: zeroruntime.MessageRoleUser, Content: "x"}}
 	called := false
 	// recover must not invoke the provider/summarize when disabled, even on a

internal/agent/completion_policy.go

@@ -1,86 +0,0 @@
-package agent
-
-// CompletionDecision is the deterministic outcome of evaluating a text-only
-// assistant turn. Uncertain means the loop needs one bounded follow-up action
-// before it can finalize the run.
-type CompletionDecision string
-
-const (
-	CompletionUncertain  CompletionDecision = "uncertain"
-	CompletionComplete   CompletionDecision = "complete"
-	CompletionIncomplete CompletionDecision = "incomplete"
-)
-
-type completionAction string
-
-const (
-	completionActionNone          completionAction = ""
-	completionActionContinue      completionAction = "continue"
-	completionActionSemanticCheck completionAction = "semantic_check"
-)
-
-type completionEvaluation struct {
-	Decision CompletionDecision
-	Action   completionAction
-	Reason   string
-}
-
-// completionPolicy owns the small amount of state needed to keep uncertain
-// decisions bounded. requireSemanticCheck is set only for run profiles that
-// already opted into self-correction; ordinary completion remains entirely
-// local and adds no model call.
-type completionPolicy struct {
-	continueNudges         int
-	requireSemanticCheck   bool
-	semanticCheckRequested bool
-}
-
-func newCompletionPolicy(requireSemanticCheck bool) *completionPolicy {
-	return &completionPolicy{requireSemanticCheck: requireSemanticCheck}
-}
-
-func (policy *completionPolicy) evaluate(text string, context completionContext) completionEvaluation {
-	// A direct admission is strong local evidence and takes precedence over all
-	// weaker signals, avoiding both continuation nudges and semantic checks.
-	if reason := selfReportedIncompletion(text); reason != "" {
-		return completionEvaluation{Decision: CompletionIncomplete, Reason: reason}
-	}
-
-	// A continuation cue is strong unfinished evidence; a pending plan is only
-	// weak evidence because bookkeeping can be stale. Both get a bounded chance
-	// to continue, but only a persisted cue is ultimately incomplete.
-	cue := endsWithContinuationCue(text)
-	if cue || context.PlanPending {
-		if policy.continueNudges < maxContinueNudges {
-			policy.continueNudges++
-			reason := "your message ended mid-step"
-			if !cue {
-				reason = "pending plan items remain — finish them, or mark them complete with update_plan if you are done"
-			}
-			return completionEvaluation{
-				Decision: CompletionUncertain,
-				Action:   completionActionContinue,
-				Reason:   reason,
-			}
-		}
-		if cue {
-			return completionEvaluation{
-				Decision: CompletionIncomplete,
-				Reason:   "your message ended mid-step",
-			}
-		}
-	}
-
-	// Profiles with self-correction enabled require one task-grounded semantic
-	// check. It is requested at most once per run; the next locally complete turn
-	// is accepted without another model call.
-	if policy.requireSemanticCheck && !policy.semanticCheckRequested {
-		policy.semanticCheckRequested = true
-		return completionEvaluation{
-			Decision: CompletionUncertain,
-			Action:   completionActionSemanticCheck,
-		}
-	}
-
-	return completionEvaluation{Decision: CompletionComplete, Action: completionActionNone}
-}

internal/agent/completion_policy_test.go

@@ -1,75 +0,0 @@
-package agent
-
-import (
-	"strings"
-	"testing"
-)
-
-func TestCompletionPolicyLocalEvidenceDecidesWithoutSemanticCheck(t *testing.T) {
-	policy := newCompletionPolicy(false)
-
-	complete := policy.evaluate("Done. All required checks pass.", completionContext{})
-	if complete.Decision != CompletionComplete {
-		t.Fatalf("confident completion decision = %q, want %q", complete.Decision, CompletionComplete)
-	}
-
-	incomplete := policy.evaluate("I couldn't verify the result, so this is my best guess.", completionContext{})
-	if incomplete.Decision != CompletionIncomplete {
-		t.Fatalf("admitted failure decision = %q, want %q", incomplete.Decision, CompletionIncomplete)
-	}
-}
-
-func TestCompletionPolicyPreservesBoundedPlanStallProtection(t *testing.T) {
-	policy := newCompletionPolicy(false)
-	for attempt := 0; attempt < maxContinueNudges; attempt++ {
-		got := policy.evaluate("Let me inspect the remaining configuration:", completionContext{PlanPending: true})
-		if got.Decision != CompletionUncertain || got.Action != completionActionContinue {
-			t.Fatalf("attempt %d = (%q, %q), want uncertain continue", attempt+1, got.Decision, got.Action)
-		}
-	}
-
-	got := policy.evaluate("Let me inspect the remaining configuration:", completionContext{PlanPending: true})
-	if got.Decision != CompletionIncomplete {
-		t.Fatalf("decision after nudge budget = %q, want %q", got.Decision, CompletionIncomplete)
-	}
-}
-
-func TestCompletionPolicyTreatsPendingPlanAsWeakEvidence(t *testing.T) {
-	policy := newCompletionPolicy(false)
-	for attempt := 0; attempt < maxContinueNudges; attempt++ {
-		got := policy.evaluate("All set.", completionContext{PlanPending: true})
-		if got.Decision != CompletionUncertain || got.Action != completionActionContinue {
-			t.Fatalf("attempt %d = (%q, %q), want uncertain continue", attempt+1, got.Decision, got.Action)
-		}
-	}
-
-	got := policy.evaluate("All set.", completionContext{PlanPending: true})
-	if got.Decision != CompletionComplete {
-		t.Fatalf("stale plan decision after nudge budget = %q, want %q", got.Decision, CompletionComplete)
-	}
-}
-
-func TestCompletionPolicyAllowsExactlyOneRequiredSemanticCheck(t *testing.T) {
-	policy := newCompletionPolicy(true)
-
-	first := policy.evaluate("Implemented and tested.", completionContext{})
-	if first.Decision != CompletionUncertain || first.Action != completionActionSemanticCheck {
-		t.Fatalf("first decision = (%q, %q), want uncertain semantic check", first.Decision, first.Action)
-	}
-
-	second := policy.evaluate("PASS. The result meets the task criterion.", completionContext{})
-	if second.Decision != CompletionComplete {
-		t.Fatalf("post-check decision = %q, want %q", second.Decision, CompletionComplete)
-	}
-	if second.Action == completionActionSemanticCheck {
-		t.Fatal("semantic check was requested more than once")
-	}
-}
-
-func TestAcceptanceVerificationNudgeIncludesBoundedObjective(t *testing.T) {
-	objective := "make the requested behavior observable"
-	nudge := acceptanceVerificationNudge(objective)
-	if !strings.Contains(nudge, "Task objective: "+objective) {
-		t.Fatalf("semantic check was not grounded in the objective: %q", nudge)
-	}
-}

internal/agent/context_measurement.go

@@ -23,24 +23,6 @@ type ContextBreakdown struct {
 	TotalTokens   int     // SystemTokens + ToolTokens + MessageTokens
 	ContextWindow int     // model context window; 0 when unknown
 	UsedFraction  float64 // TotalTokens / ContextWindow; 0 when window unknown
-	// Blocks explain why each model-visible category is present without exposing
-	// its content. The planner currently preserves all three categories.
-	Blocks []ContextBlock
-	// CompletePrefixHash identifies the provider-visible system+tool prefix.
-	CompletePrefixHash string
-	// PrefixInvalidationReason is "initial", "unchanged", or a stable comma-
-	// separated list of changed prefix categories.
-	PrefixInvalidationReason string
-}
-
-// ContextBlock is content-free evidence for one model-visible context category.
-type ContextBlock struct {
-	Kind        string
-	Tokens      int
-	CacheClass  string
-	Authority   string
-	Recoverable bool
-	Reason      string
 }
 
 // MeasureContext estimates the per-category token footprint of a request: the
@@ -59,24 +41,6 @@ func MeasureContext(messages []zeroruntime.Message, tools []zeroruntime.ToolDefi
 		ContextWindow: contextWindow,
 	}
 	breakdown.TotalTokens = breakdown.SystemTokens + breakdown.ToolTokens + breakdown.MessageTokens
-	if breakdown.SystemTokens > 0 {
-		breakdown.Blocks = append(breakdown.Blocks, ContextBlock{
-			Kind: "system", Tokens: breakdown.SystemTokens, CacheClass: "stable_prefix",
-			Authority: "required", Reason: "instructions, workspace policy, and safety policy",
-		})
-	}
-	if breakdown.ToolTokens > 0 {
-		breakdown.Blocks = append(breakdown.Blocks, ContextBlock{
-			Kind: "tools", Tokens: breakdown.ToolTokens, CacheClass: "stable_prefix",
-			Authority: "capability", Reason: "currently permitted and selected tool definitions",
-		})
-	}
-	if breakdown.MessageTokens > 0 {
-		breakdown.Blocks = append(breakdown.Blocks, ContextBlock{
-			Kind: "conversation", Tokens: breakdown.MessageTokens, CacheClass: "dynamic_tail",
-			Authority: "transcript", Reason: "valid provider replay history and current task context",
-		})
-	}
 	if contextWindow > 0 {
 		breakdown.UsedFraction = float64(breakdown.TotalTokens) / float64(contextWindow)
 	}

internal/agent/context_measurement_test.go

@@ -32,9 +32,6 @@ func TestMeasureContextSplitsByCategory(t *testing.T) {
 	if breakdown.UsedFraction != wantFraction {
 		t.Fatalf("UsedFraction = %v, want %v", breakdown.UsedFraction, wantFraction)
 	}
-	if len(breakdown.Blocks) != 3 {
-		t.Fatalf("Blocks = %#v, want system/tools/conversation", breakdown.Blocks)
-	}
 }
 
 func TestMeasureContextUnknownWindowHasZeroFraction(t *testing.T) {

internal/agent/context_planner.go

@@ -1,185 +0,0 @@
-package agent
-
-import (
-	"strings"
-
-	"github.com/Gitlawb/zero/internal/zeroruntime"
-)
-
-type contextPlannerConfig struct {
-	contextWindow  int
-	promptCacheKey string
-	serviceTier    string
-	promptParts    systemPromptParts
-}
-
-// contextPlanner is the single seam for constructing model-visible requests.
-// Planning is deterministic and does not retrieve content, execute tools, or
-// change permissions. The initial implementation deliberately preserves every
-// message and tool definition while making composition and cache drift
-// inspectable; later selection policies must continue to cross this seam.
-type contextPlanner struct {
-	config          contextPlannerConfig
-	previousPrefix  prefixFingerprint
-	hasPrevious     bool
-	toolSnapshotKey string
-	toolSnapshot    []zeroruntime.ToolDefinition
-}
-
-type contextPlan struct {
-	Request           zeroruntime.CompletionRequest
-	Breakdown         ContextBreakdown
-	PrefixFingerprint prefixFingerprint
-}
-
-func newContextPlanner(config contextPlannerConfig) *contextPlanner {
-	return &contextPlanner{config: config}
-}
-
-// Plan returns a provider request snapshot plus content-free accounting.
-// It intentionally performs no relevance filtering: preserving current model
-// capability is the baseline contract for future planner policies.
-func (planner *contextPlanner) Plan(messages []zeroruntime.Message, toolDefs []zeroruntime.ToolDefinition, reasoningEffort string) contextPlan {
-	return planner.plan(messages, toolDefs, reasoningEffort, planner.config.promptParts)
-}
-
-// planWithPromptParts plans a request whose stable system sections differ from
-// the planner's configured main-run prompt, such as a compaction summary call.
-func (planner *contextPlanner) planWithPromptParts(messages []zeroruntime.Message, toolDefs []zeroruntime.ToolDefinition, reasoningEffort string, parts systemPromptParts) contextPlan {
-	return planner.plan(messages, toolDefs, reasoningEffort, parts)
-}
-
-func (planner *contextPlanner) plan(messages []zeroruntime.Message, toolDefs []zeroruntime.ToolDefinition, reasoningEffort string, parts systemPromptParts) contextPlan {
-	requestMessages := copyMessages(messages)
-	requestSystemPrompt := leadingSystemContent(requestMessages)
-	if parts.prompt != requestSystemPrompt {
-		// Component boundaries from a different request would make the detailed
-		// invalidation reason misleading. Retain the exact request prefix and
-		// classify it conservatively as an unstructured system prompt instead.
-		parts = systemPromptParts{prompt: requestSystemPrompt}
-	}
-	fingerprint := computePrefixFingerprint(buildPromptSubstringsFromParts(parts, toolDefs))
-	request := zeroruntime.CompletionRequest{
-		Messages:        requestMessages,
-		Tools:           planner.snapshotTools(toolDefs, fingerprint),
-		ReasoningEffort: reasoningEffort,
-		ServiceTier:     planner.config.serviceTier,
-		PromptCacheKey:  planner.config.promptCacheKey,
-	}
-	breakdown := MeasureContext(request.Messages, request.Tools, planner.config.contextWindow)
-	breakdown.CompletePrefixHash = fingerprint.CompletePrefixHash
-	var previous *prefixFingerprint
-	if planner.hasPrevious {
-		previous = &planner.previousPrefix
-	}
-	breakdown.PrefixInvalidationReason = explainPrefixChange(previous, fingerprint)
-	planner.previousPrefix = fingerprint
-	planner.hasPrevious = true
-	return contextPlan{
-		Request:           request,
-		Breakdown:         breakdown,
-		PrefixFingerprint: fingerprint,
-	}
-}
-
-// snapshotTools freezes provider-visible tool definitions when their semantic
-// fingerprint changes. Stable turns reuse the frozen schemas instead of
-// recursively cloning every map and slice for every request.
-func (planner *contextPlanner) snapshotTools(toolDefs []zeroruntime.ToolDefinition, fingerprint prefixFingerprint) []zeroruntime.ToolDefinition {
-	key := fingerprint.ToolsHash + "\x00" + fingerprint.SchemaHash
-	if planner.toolSnapshotKey != key {
-		planner.toolSnapshot = copyToolDefinitions(toolDefs)
-		planner.toolSnapshotKey = key
-	}
-	return planner.toolSnapshot
-}
-
-func copyToolDefinitions(toolDefs []zeroruntime.ToolDefinition) []zeroruntime.ToolDefinition {
-	if toolDefs == nil {
-		return nil
-	}
-	copied := make([]zeroruntime.ToolDefinition, len(toolDefs))
-	for index, definition := range toolDefs {
-		copied[index] = definition
-		copied[index].Parameters = copySchemaMap(definition.Parameters)
-	}
-	return copied
-}
-
-func copySchemaMap(schema map[string]any) map[string]any {
-	if schema == nil {
-		return nil
-	}
-	copied := make(map[string]any, len(schema))
-	for key, value := range schema {
-		copied[key] = copySchemaValue(value)
-	}
-	return copied
-}
-
-func copySchemaValue(value any) any {
-	switch typed := value.(type) {
-	case map[string]any:
-		return copySchemaMap(typed)
-	case []any:
-		copied := make([]any, len(typed))
-		for index, item := range typed {
-			copied[index] = copySchemaValue(item)
-		}
-		return copied
-	case []string:
-		return append([]string(nil), typed...)
-	default:
-		return value
-	}
-}
-
-func leadingSystemContent(messages []zeroruntime.Message) string {
-	contents := make([]string, 0, 1)
-	for _, message := range messages {
-		if message.Role != zeroruntime.MessageRoleSystem {
-			break
-		}
-		contents = append(contents, message.Content)
-	}
-	return strings.Join(contents, "\n\n")
-}
-
-func explainPrefixChange(previous *prefixFingerprint, current prefixFingerprint) string {
-	if previous == nil {
-		return "initial"
-	}
-	if previous.CompletePrefixHash == current.CompletePrefixHash {
-		return "unchanged"
-	}
-	reasons := make([]string, 0, 3)
-	if previous.SystemPromptHash != current.SystemPromptHash {
-		parts := make([]string, 0, 4)
-		if previous.BaseInstructionsHash != current.BaseInstructionsHash {
-			parts = append(parts, "base_instructions")
-		}
-		if previous.ConfirmationPolicyHash != current.ConfirmationPolicyHash {
-			parts = append(parts, "confirmation_policy")
-		}
-		if previous.ProjectContextHash != current.ProjectContextHash {
-			parts = append(parts, "project_context")
-		}
-		if previous.SkillsHash != current.SkillsHash {
-			parts = append(parts, "skills")
-		}
-		if len(parts) == 0 {
-			parts = append(parts, "system_prompt")
-		}
-		reasons = append(reasons, parts...)
-	}
-	if previous.ToolsHash != current.ToolsHash {
-		reasons = append(reasons, "tools")
-	}
-	if previous.SchemaHash != current.SchemaHash {
-		reasons = append(reasons, "schema")
-	}
-	if len(reasons) == 0 {
-		return "prefix_changed"
-	}
-	return strings.Join(reasons, ",")
-}

internal/agent/context_planner_test.go

@@ -1,187 +0,0 @@
-package agent
-
-import (
-	"reflect"
-	"testing"
-
-	"github.com/Gitlawb/zero/internal/zeroruntime"
-)
-
-func TestContextPlannerPreservesProviderRequest(t *testing.T) {
-	messages := []zeroruntime.Message{
-		{Role: zeroruntime.MessageRoleSystem, Content: "system"},
-		{Role: zeroruntime.MessageRoleUser, Content: "inspect this", Images: []zeroruntime.ImageBlock{{MediaType: "image/png", Data: []byte{1, 2, 3}}}},
-		{Role: zeroruntime.MessageRoleAssistant, Content: "working", ToolCalls: []zeroruntime.ToolCall{{ID: "call-1", Name: "read_file", Arguments: `{"path":"main.go"}`}}, Reasoning: []zeroruntime.ReasoningBlock{{Provider: "test", Type: "thinking", Signature: "sig"}}},
-	}
-	toolDefs := []zeroruntime.ToolDefinition{{
-		Name: "read_file", Description: "Read a file",
-		Parameters: map[string]any{
-			"type":       "object",
-			"required":   []string{"path"},
-			"properties": map[string]any{"path": map[string]any{"type": "string"}},
-			"anyOf":      []any{map[string]any{"type": "object"}},
-		},
-	}}
-	planner := newContextPlanner(contextPlannerConfig{
-		contextWindow:  128_000,
-		promptCacheKey: "session-1",
-		serviceTier:    "priority",
-		promptParts:    systemPromptParts{prompt: "system", baseInstructions: "system"},
-	})
-
-	plan := planner.Plan(messages, toolDefs, "medium")
-	want := zeroruntime.CompletionRequest{
-		Messages:        copyMessages(messages),
-		Tools:           toolDefs,
-		ReasoningEffort: "medium",
-		ServiceTier:     "priority",
-		PromptCacheKey:  "session-1",
-	}
-	if !reflect.DeepEqual(plan.Request, want) {
-		t.Fatalf("planned request changed provider input:\n got: %#v\nwant: %#v", plan.Request, want)
-	}
-	stable := planner.Plan(messages, toolDefs, "medium")
-	if &stable.Request.Tools[0] != &plan.Request.Tools[0] {
-		t.Fatal("stable tool fingerprint rebuilt the immutable schema snapshot")
-	}
-
-	// A plan owns its request snapshot; later caller mutation cannot change its
-	// messages or its recursively nested tool schema.
-	messages[1].Content = "changed"
-	messages[1].Images[0].Data[0] = 9
-	toolDefs[0].Name = "changed_tool"
-	toolDefs[0].Parameters["required"].([]string)[0] = "other"
-	toolDefs[0].Parameters["properties"].(map[string]any)["path"].(map[string]any)["type"] = "integer"
-	toolDefs[0].Parameters["anyOf"].([]any)[0].(map[string]any)["type"] = "array"
-	if plan.Request.Messages[1].Content != "inspect this" || plan.Request.Messages[1].Images[0].Data[0] != 1 {
-		t.Fatalf("planned messages alias caller state: %#v", plan.Request.Messages[1])
-	}
-	if plan.Request.Tools[0].Name != "read_file" ||
-		plan.Request.Tools[0].Parameters["required"].([]string)[0] != "path" ||
-		plan.Request.Tools[0].Parameters["properties"].(map[string]any)["path"].(map[string]any)["type"] != "string" ||
-		plan.Request.Tools[0].Parameters["anyOf"].([]any)[0].(map[string]any)["type"] != "object" {
-		t.Fatalf("planned tools alias caller state: %#v", plan.Request.Tools[0])
-	}
-
-	next := planner.Plan(messages, toolDefs, "medium")
-	if next.PrefixFingerprint == plan.PrefixFingerprint || next.Request.Tools[0].Name != "changed_tool" {
-		t.Fatalf("changed tool schema did not refresh snapshot: %#v", next)
-	}
-	if plan.Request.Tools[0].Name != "read_file" {
-		t.Fatalf("refresh mutated earlier request: %#v", plan.Request.Tools[0])
-	}
-}
-
-func TestContextPlannerReportsInspectableBlocks(t *testing.T) {
-	planner := newContextPlanner(contextPlannerConfig{contextWindow: 100_000})
-	plan := planner.Plan([]zeroruntime.Message{
-		{Role: zeroruntime.MessageRoleSystem, Content: filler(400)},
-		{Role: zeroruntime.MessageRoleUser, Content: filler(200)},
-	}, []zeroruntime.ToolDefinition{{Name: "read_file", Description: "Read a file"}}, "")
-
-	if plan.Breakdown.PrefixInvalidationReason != "initial" || plan.Breakdown.CompletePrefixHash == "" {
-		t.Fatalf("prefix evidence = %#v", plan.Breakdown)
-	}
-	if len(plan.Breakdown.Blocks) != 3 {
-		t.Fatalf("blocks = %#v, want system/tools/conversation", plan.Breakdown.Blocks)
-	}
-	wantKinds := []string{"system", "tools", "conversation"}
-	total := 0
-	for index, block := range plan.Breakdown.Blocks {
-		if block.Kind != wantKinds[index] || block.Tokens <= 0 || block.Reason == "" || block.Authority == "" || block.CacheClass == "" {
-			t.Fatalf("block %d = %#v", index, block)
-		}
-		total += block.Tokens
-	}
-	if total != plan.Breakdown.TotalTokens {
-		t.Fatalf("block tokens = %d, want total %d", total, plan.Breakdown.TotalTokens)
-	}
-}
-
-func TestContextPlannerFingerprintsActualRequestSystemPrompt(t *testing.T) {
-	planner := newContextPlanner(contextPlannerConfig{promptParts: systemPromptParts{
-		prompt: "configured system", baseInstructions: "configured system",
-	}})
-	messages := []zeroruntime.Message{
-		{Role: zeroruntime.MessageRoleSystem, Content: "request-specific system"},
-		{Role: zeroruntime.MessageRoleUser, Content: "task"},
-	}
-
-	plan := planner.Plan(messages, nil, "")
-	want := computePrefixFingerprint(buildPromptSubstringsFromParts(
-		systemPromptParts{prompt: "request-specific system"}, nil,
-	))
-	if plan.PrefixFingerprint != want {
-		t.Fatalf("fingerprint describes configured prompt instead of request:\n got: %#v\nwant: %#v", plan.PrefixFingerprint, want)
-	}
-}
-
-func TestContextPlannerExplainsPrefixInvalidation(t *testing.T) {
-	planner := newContextPlanner(contextPlannerConfig{promptParts: systemPromptParts{
-		prompt: "system", baseInstructions: "base", projectContext: "project",
-	}})
-	messages := []zeroruntime.Message{{Role: zeroruntime.MessageRoleSystem, Content: "system"}, {Role: zeroruntime.MessageRoleUser, Content: "task"}}
-	tools := []zeroruntime.ToolDefinition{{Name: "read_file", Description: "Read", Parameters: map[string]any{"type": "object"}}}
-
-	if got := planner.Plan(messages, tools, "").Breakdown.PrefixInvalidationReason; got != "initial" {
-		t.Fatalf("first reason = %q, want initial", got)
-	}
-	if got := planner.Plan(messages, tools, "").Breakdown.PrefixInvalidationReason; got != "unchanged" {
-		t.Fatalf("stable reason = %q, want unchanged", got)
-	}
-	tools[0].Description = "Read an exact file"
-	if got := planner.Plan(messages, tools, "").Breakdown.PrefixInvalidationReason; got != "tools" {
-		t.Fatalf("tool reason = %q, want tools", got)
-	}
-	tools[0].Parameters["required"] = []any{"path"}
-	if got := planner.Plan(messages, tools, "").Breakdown.PrefixInvalidationReason; got != "schema" {
-		t.Fatalf("schema reason = %q, want schema", got)
-	}
-	planner.config.promptParts.prompt = "system with changed project"
-	planner.config.promptParts.projectContext = "changed project"
-	messages[0].Content = "system with changed project"
-	if got := planner.Plan(messages, tools, "").Breakdown.PrefixInvalidationReason; got != "project_context" {
-		t.Fatalf("project reason = %q, want project_context", got)
-	}
-}
-
-func TestExplainPrefixChangeNamesEveryChangedComponent(t *testing.T) {
-	baseline := prefixFingerprint{
-		SystemPromptHash:       "system",
-		BaseInstructionsHash:   "base",
-		ConfirmationPolicyHash: "policy",
-		ProjectContextHash:     "project",
-		SkillsHash:             "skills",
-		ToolsHash:              "tools",
-		SchemaHash:             "schema",
-		CompletePrefixHash:     "complete",
-	}
-	tests := []struct {
-		name   string
-		mutate func(*prefixFingerprint)
-		want   string
-	}{
-		{name: "base instructions", mutate: func(value *prefixFingerprint) {
-			value.SystemPromptHash, value.BaseInstructionsHash = "changed", "changed"
-		}, want: "base_instructions"},
-		{name: "confirmation policy", mutate: func(value *prefixFingerprint) {
-			value.SystemPromptHash, value.ConfirmationPolicyHash = "changed", "changed"
-		}, want: "confirmation_policy"},
-		{name: "project context", mutate: func(value *prefixFingerprint) {
-			value.SystemPromptHash, value.ProjectContextHash = "changed", "changed"
-		}, want: "project_context"},
-		{name: "skills", mutate: func(value *prefixFingerprint) { value.SystemPromptHash, value.SkillsHash = "changed", "changed" }, want: "skills"},
-		{name: "unclassified system prompt", mutate: func(value *prefixFingerprint) { value.SystemPromptHash = "changed" }, want: "system_prompt"},
-		{name: "tools and schema", mutate: func(value *prefixFingerprint) { value.ToolsHash, value.SchemaHash = "changed", "changed" }, want: "tools,schema"},
-	}
-	for _, test := range tests {
-		t.Run(test.name, func(t *testing.T) {
-			current := baseline
-			test.mutate(&current)
-			current.CompletePrefixHash = "changed"
-			if got := explainPrefixChange(&baseline, current); got != test.want {
-				t.Fatalf("reason = %q, want %q", got, test.want)
-			}
-		})
-	}
-}

internal/agent/deferred_loop_test.go

@@ -150,7 +150,7 @@ func assertNoDeferredDiscoveryMessage(t *testing.T, request zeroruntime.Completi
 func TestPartitionToolsInactiveIsByteIdenticalAndDropsToolSearch(t *testing.T) {
 	root := t.TempDir()
 	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedReadFileTool(root, nil))
+	registry.Register(tools.NewReadFileTool(root))
 	registry.Register(fakeDeferredTool{name: "mcp__srv__a", desc: "tool a"})
 	registry.Register(fakeToolSearchTool{})
 
@@ -305,8 +305,8 @@ func TestPartitionToolsActiveExcludesDisabledDeferredFromDiscoveryAndExposed(t *
 func TestPartitionToolsActiveHidesUnloadedExposesLoaded(t *testing.T) {
 	root := t.TempDir()
 	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedReadFileTool(root, nil)) // non-deferred builtin
-	registry.Register(fakeToolSearchTool{})                   // non-deferred, must stay exposed
+	registry.Register(tools.NewReadFileTool(root)) // non-deferred builtin
+	registry.Register(fakeToolSearchTool{})        // non-deferred, must stay exposed
 	registry.Register(fakeDeferredTool{name: "mcp__srv__alpha", desc: "alpha tool"})
 	registry.Register(fakeDeferredTool{name: "mcp__srv__beta", desc: "beta tool"})
 
@@ -341,15 +341,15 @@ func TestPartitionToolsActiveHidesUnloadedExposesLoaded(t *testing.T) {
 	if search.Description != discovery || !strings.Contains(search.Description, "mcp") {
 		t.Fatalf("tool_search description must carry discovery source, got %q", search.Description)
 	}
-	if !strings.Contains(search.Description, "mcp__srv__alpha") {
-		t.Fatalf("stable tool_search catalog must retain already-loaded tool names, got %q", search.Description)
+	if strings.Contains(search.Description, "mcp__srv__alpha") {
+		t.Fatalf("tool_search description must not list already-loaded tool names, got %q", search.Description)
 	}
 	if !strings.Contains(search.Description, "mcp__srv__beta") {
 		t.Fatalf("tool_search description must list exact hidden tool names, got %q", search.Description)
 	}
 }
 
-func TestPartitionToolsActiveNothingHiddenKeepsStableCatalog(t *testing.T) {
+func TestPartitionToolsActiveNothingHiddenEmptyReminder(t *testing.T) {
 	registry := tools.NewRegistry()
 	registry.Register(fakeDeferredTool{name: "mcp__srv__alpha", desc: "alpha"})
 	registry.Register(fakeDeferredTool{name: "mcp__srv__beta", desc: "beta"})
@@ -368,8 +368,9 @@ func TestPartitionToolsActiveNothingHiddenKeepsStableCatalog(t *testing.T) {
 	if !exposedNames["tool_search"] {
 		t.Fatalf("expected tool_search exposed on active path, got %#v", exposed)
 	}
-	if !strings.Contains(reminder, "mcp__srv__alpha") || !strings.Contains(reminder, "mcp__srv__beta") {
-		t.Fatalf("stable catalog must retain all deferred names after loading, got %q", reminder)
+	// No hidden tools means no dynamic discovery text is needed.
+	if reminder != "" {
+		t.Fatalf("expected empty discovery text when nothing is hidden, got %q", reminder)
 	}
 }
 
@@ -797,98 +798,3 @@ func TestDisabledToolSearchFallsBackToEager(t *testing.T) {
 		t.Fatalf("deferred tool must be callable under eager fallback, got status=%s output=%q", result.Status, result.Output)
 	}
 }
-
-// fakeDeferredMutatorTool is a deferred-eligible tool with mutating Safety
-// (SideEffectWrite), standing in for a real write/mutator MCP tool that would
-// be hidden behind tool_search once deferral activates.
-type fakeDeferredMutatorTool struct{ name string }
-
-func (t fakeDeferredMutatorTool) Name() string        { return t.name }
-func (t fakeDeferredMutatorTool) Description() string { return "mutates the workspace, deferred" }
-func (t fakeDeferredMutatorTool) Parameters() tools.Schema {
-	return tools.Schema{Type: "object", AdditionalProperties: false}
-}
-func (t fakeDeferredMutatorTool) Safety() tools.Safety {
-	return tools.Safety{SideEffect: tools.SideEffectWrite, Permission: tools.PermissionPrompt, Reason: "mutates"}
-}
-func (t fakeDeferredMutatorTool) Run(_ context.Context, _ map[string]any) tools.Result {
-	return tools.Result{Status: tools.StatusOK, Output: "mutated"}
-}
-func (t fakeDeferredMutatorTool) Deferred() bool { return true }
-
-// TestPlanModeToolSearchNeverLeaksDeferredMutatorSchema guards the concern
-// raised in PR #642 review (jatmn, P2): that tool_search filters deferred
-// candidates only by EnabledTools/DisabledTools, not by plan-mode visibility,
-// so a plan-mode model could call `tool_search select:<deferred write tool>`
-// and receive that tool's full schema even though a direct call to it is
-// correctly denied the following turn.
-//
-// tool_search's own Safety is SideEffectNone, and ToolAdvertisedForPermissionMode (the
-// same gate executeToolCall uses to deny a direct call) requires
-// SideEffect==Read to advertise a tool in plan mode. That means tool_search
-// itself is never advertised, never activates deferral (loaderUsable in
-// partitionToolsCached requires ToolAdvertised(loader, permissionMode)), and
-// is denied at dispatch like any other hidden tool if a stale/adversarial
-// call reaches it anyway — so a deferred mutator's schema can never reach the
-// model through tool_search while in plan mode. This test pins that
-// end-to-end: tool_search is absent from the advertised tool list, and a
-// forced call to it is denied before rendering any tool schema.
-func TestPlanModeToolSearchNeverLeaksDeferredMutatorSchema(t *testing.T) {
-	root := t.TempDir()
-	registry := tools.NewRegistry()
-	registry.Register(tools.NewReadFileTool(root))
-	registry.Register(fakeDeferredMutatorTool{name: "mcp__srv__mutate"})
-	registry.Register(fakeDeferredMutatorTool{name: "mcp__srv__mutate2"})
-	registry.Register(tools.NewToolSearchTool(registry))
-
-	provider := &mockProvider{turns: [][]zeroruntime.StreamEvent{
-		{ // turn 1: force a call to tool_search even though it should not be advertised.
-			{Type: zeroruntime.StreamEventToolCallStart, ToolCallID: "c1", ToolName: tools.ToolSearchToolName},
-			{Type: zeroruntime.StreamEventToolCallDelta, ToolCallID: "c1", ArgumentsFragment: `{"query":"select:mcp__srv__mutate"}`},
-			{Type: zeroruntime.StreamEventToolCallEnd, ToolCallID: "c1"},
-			{Type: zeroruntime.StreamEventDone},
-		},
-		{ // turn 2: final answer.
-			{Type: zeroruntime.StreamEventText, Content: "done"},
-			{Type: zeroruntime.StreamEventDone},
-		},
-	}}
-
-	result, err := Run(context.Background(), "plan", provider, Options{
-		Registry:       registry,
-		PermissionMode: PermissionModePlan,
-		DeferThreshold: 2, // 2 deferred mutators registered => eligible for deferral.
-		MaxTurns:       2,
-	})
-	if err != nil {
-		t.Fatal(err)
-	}
-
-	// tool_search (and the deferred mutators) must not be advertised in plan
-	// mode's turn 1 tool list at all.
-	for _, def := range provider.requests[0].Tools {
-		if def.Name == tools.ToolSearchToolName {
-			t.Fatalf("plan mode must not advertise tool_search, got %#v", provider.requests[0].Tools)
-		}
-		if def.Name == "mcp__srv__mutate" || def.Name == "mcp__srv__mutate2" {
-			t.Fatalf("plan mode must not advertise a deferred mutator, got %#v", provider.requests[0].Tools)
-		}
-	}
-
-	// The forced call must be denied outright, never a loaded-schema result:
-	// the tool result must not mention the mutator's name or carry a
-	// load_tools signal.
-	var toolMessage string
-	for _, message := range result.Messages {
-		if message.Role == zeroruntime.MessageRoleTool {
-			toolMessage = message.Content
-			break
-		}
-	}
-	if !strings.Contains(toolMessage, "not available in plan mode") {
-		t.Fatalf("expected tool_search call denied in plan mode, got %q", toolMessage)
-	}
-	if strings.Contains(toolMessage, "mcp__srv__mutate") {
-		t.Fatalf("denial must not leak the deferred mutator's name/schema, got %q", toolMessage)
-	}
-}

internal/agent/denial_test.go

@@ -11,7 +11,7 @@ import (
 func TestExecuteToolCallCategorizesFilteredDenial(t *testing.T) {
 	root := t.TempDir()
 	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedReadFileTool(root, nil))
+	registry.Register(tools.NewReadFileTool(root))
 
 	result, _ := executeToolCall(context.Background(), registry, ToolCall{ID: "c1", Name: "read_file", Arguments: `{"path":"x"}`}, PermissionModeAuto, Options{
 		DisabledTools: []string{"read_file"},

internal/agent/escalate_loop_test.go

@@ -144,10 +144,9 @@ func TestRunSwitchesProviderOnEscalationRequest(t *testing.T) {
 	var switchedTo string
 	switchCount := 0
 	result, err := Run(context.Background(), "go", firstProvider, Options{
-		Registry:    registry,
-		Model:       "claude-sonnet-4.5",
-		ServiceTier: "priority",
-		MaxTurns:    4,
+		Registry: registry,
+		Model:    "claude-sonnet-4.5",
+		MaxTurns: 4,
 		ModelSwitcher: func(_ context.Context, modelID string) (Provider, error) {
 			switchCount++
 			switchedTo = modelID
@@ -168,15 +167,9 @@ func TestRunSwitchesProviderOnEscalationRequest(t *testing.T) {
 	if len(firstProvider.requests) != 1 {
 		t.Fatalf("expected first provider to handle exactly the escalation turn, got %d requests", len(firstProvider.requests))
 	}
-	if got := firstProvider.requests[0].ServiceTier; got != "priority" {
-		t.Fatalf("initial request service tier = %q, want priority", got)
-	}
 	if len(secondProvider.requests) != 1 {
 		t.Fatalf("expected second provider to handle the post-switch turn, got %d requests", len(secondProvider.requests))
 	}
-	if got := secondProvider.requests[0].ServiceTier; got != "" {
-		t.Fatalf("post-switch request retained service tier %q", got)
-	}
 	if result.FinalAnswer != "answered on the upgraded model" {
 		t.Fatalf("expected final answer from the swapped provider, got %q", result.FinalAnswer)
 	}
@@ -218,10 +211,9 @@ func TestRunEscalationSwitcherErrorIsNonFatal(t *testing.T) {
 	provider := &mockProvider{turns: escalateThenAnswerTurns("recovered")}
 
 	result, err := Run(context.Background(), "go", provider, Options{
-		Registry:    registry,
-		Model:       "claude-sonnet-4.5",
-		ServiceTier: "priority",
-		MaxTurns:    4,
+		Registry: registry,
+		Model:    "claude-sonnet-4.5",
+		MaxTurns: 4,
 		ModelSwitcher: func(_ context.Context, _ string) (Provider, error) {
 			return nil, errors.New("provider build blew up")
 		},
@@ -233,9 +225,6 @@ func TestRunEscalationSwitcherErrorIsNonFatal(t *testing.T) {
 	if len(provider.requests) != 2 {
 		t.Fatalf("expected the run to continue on the original provider, got %d requests", len(provider.requests))
 	}
-	if got := provider.requests[1].ServiceTier; got != "priority" {
-		t.Fatalf("failed-switch request service tier = %q, want priority", got)
-	}
 	if result.FinalAnswer != "recovered" {
 		t.Fatalf("expected final answer after non-fatal switch error, got %q", result.FinalAnswer)
 	}

internal/agent/export_test.go

@@ -1,41 +0,0 @@
-// Test seams: helpers only test code uses, kept out of the production binary.
-package agent
-
-import (
-	"github.com/Gitlawb/zero/internal/tools"
-	"github.com/Gitlawb/zero/internal/zeroruntime"
-)
-
-func equalStringSlices(left []string, right []string) bool {
-	if len(left) != len(right) {
-		return false
-	}
-	for index := range left {
-		if left[index] != right[index] {
-			return false
-		}
-	}
-	return true
-}
-
-// parsePreservedState recovers the plan + skills from a prior summary's preserved
-// block. JSON escaping makes this lossless even when a skill body contains
-// markdown headings, code fences, or quotes. Returns ("", nil) when absent or
-// malformed.
-func parsePreservedState(summaryContent string) (string, []skillEntry) {
-	state := parsePreservedStateBlock(summaryContent)
-	return state.Plan, preservedSkillsToEntries(state.Skills)
-}
-
-// partitionTools builds the per-turn advertised tool list and optional
-// tool_search discovery text. INACTIVE (DeferThreshold <= 0 or the eligible count is
-// below it): every visible tool is exposed with its full schema EXCEPT tool_search
-// (dropped so it is never advertised when it cannot help), and the discovery text is
-// empty — byte-identical to the pre-deferral output. ACTIVE: a deferred-eligible
-// tool is exposed only when loaded[name]; otherwise it is hidden and searchable
-// through tool_search. Non-deferred tools (including tool_search) are always
-// exposed. The exposed slice is alpha-sorted by name, matching the legacy order
-// so the inactive path is stable.
-func partitionTools(registry *tools.Registry, permissionMode PermissionMode, options Options, loaded map[string]bool) ([]zeroruntime.ToolDefinition, string) {
-	return partitionToolsCached(registry, permissionMode, options, loaded, nil)
-}

internal/agent/guardrails.go

@@ -1,6 +1,7 @@
 package agent
 
 import (
+	"encoding/json"
 	"strconv"
 	"strings"
 
@@ -126,8 +127,9 @@ func endsWithContinuationCue(text string) bool {
 // which the update_plan tool coerces to "pending") counts as remaining, matching
 // that tool's own status normalization.
 func planStatusRemaining(status string) bool {
-	switch normalizeTaskPlanStatus(status) {
-	case "completed", "failed":
+	switch strings.ToLower(strings.TrimSpace(status)) {
+	case "completed", "complete", "done", "finished", "resolved", "✓", "x", "[x]",
+		"failed", "fail", "error", "errored", "blocked", "cancelled", "canceled", "abandoned", "skipped":
 		return false
 	default:
 		return true
@@ -181,112 +183,35 @@ var successNegationTails = []string{
 	"reproduce", "spot any", "locate any",
 }
 
-// narrativeMarkers flag a sentence as RETELLING a past exchange rather than
-// reporting the outcome of the current objective, so an inability admission in
-// that sentence is about THEN, not NOW. Grounded in a real false positive: a
-// conversational recap ("You asked if I could work autonomously … I was honest
-// that my sandbox had no repo … so I couldn't actually do it at the time") was
-// downgraded to INCOMPLETE on the quoted "i couldn't " stem even though the
-// current turn's objective (summarize where we left off) was fully met. Every
-// marker is second-person or explicitly past-referential, so a first-person
-// admission about the current task ("I couldn't complete the refactor") is
-// never masked.
-var narrativeMarkers = []string{
-	"you asked", "you said", "you wanted", "you mentioned",
-	"we discussed", "we talked", "we covered",
-	"when we last", "last time", "last session", "previous session",
-	"previous conversation", "earlier session", "earlier conversation",
-	"at the time", "back then",
-}
-
-// stripQuoted removes spans enclosed in double quotes (straight or curly) or
-// backticks, so an admission the model merely QUOTES — its own earlier message,
-// a log line, an error string — cannot fire the detector. Only BALANCED spans
-// are removed: an opening delimiter that never closes is kept as literal text,
-// so a stray quote cannot swallow the rest of the message (and with it a
-// genuine admission the detector must see). Single quotes are left alone: they
-// are overwhelmingly apostrophes ("couldn't"), and treating them as quote
-// delimiters would swallow the text between two contractions.
-func stripQuoted(s string) string {
-	var b strings.Builder
-	var span strings.Builder // pending text since the open delimiter, kept if it never closes
-	open := rune(0)
-	for _, r := range s {
-		switch {
-		case open != 0:
-			if (open == '"' && r == '"') || (open == '“' && r == '”') || (open == '`' && r == '`') {
-				open = 0
-				span.Reset()
-				continue
-			}
-			span.WriteRune(r)
-		case r == '"' || r == '“' || r == '`':
-			open = r
-			span.WriteRune(r)
-		default:
-			b.WriteRune(r)
-		}
-	}
-	b.WriteString(span.String()) // dangling delimiter: restore the span verbatim
-	return b.String()
-}
-
-// admissionSentences splits lowered text into sentence-ish fragments so the
-// detector judges each claim in its own context. Newlines split too (markdown
-// bullets are separate claims); the exact boundaries only need to keep an
-// admission next to its own narrative/negation context, not be grammatical.
-func admissionSentences(lower string) []string {
-	return strings.FieldsFunc(lower, func(r rune) bool {
-		return r == '.' || r == '!' || r == '?' || r == '\n'
-	})
-}
-
 // selfReportedIncompletion returns a short reason when the model's final text
 // admits it guessed or could not meet the objective, else "". Case-insensitive.
-// Matching is per sentence, after dropping quoted spans, and a sentence that
-// retells a past exchange (narrativeMarkers) is skipped entirely — an admission
-// must be the model's own report about the CURRENT objective, not general
-// language that merely resembles one.
 func selfReportedIncompletion(text string) string {
-	for _, sentence := range admissionSentences(strings.ToLower(stripQuoted(text))) {
-		if containsAny(sentence, narrativeMarkers) {
-			continue
+	lower := strings.ToLower(text)
+	for _, phrase := range selfReportPhrases {
+		if strings.Contains(lower, phrase) {
+			return selfReportReason(phrase)
 		}
-		for _, phrase := range selfReportPhrases {
-			if strings.Contains(sentence, phrase) {
-				return selfReportReason(phrase)
+	}
+	for _, stem := range inabilityStems {
+		// Scan EVERY occurrence of the stem, not just the first: an earlier
+		// success-negation use ("I could not find any examples, so I could not
+		// implement it") must not mask a later genuine admission with the same stem.
+		for start := 0; ; {
+			rel := strings.Index(lower[start:], stem)
+			if rel < 0 {
+				break
 			}
-		}
-		for _, stem := range inabilityStems {
-			// Scan EVERY occurrence of the stem, not just the first: an earlier
-			// success-negation use ("I could not find any examples, so I could not
-			// implement it") must not mask a later genuine admission with the same stem.
-			for start := 0; ; {
-				rel := strings.Index(sentence[start:], stem)
-				if rel < 0 {
-					break
-				}
-				abs := start + rel
-				tail := strings.TrimSpace(sentence[abs+len(stem):])
-				if !hasAnyPrefix(tail, successNegationTails) {
-					return selfReportReason(strings.TrimSpace(stem) + " …")
-				}
-				start = abs + len(stem)
+			abs := start + rel
+			tail := strings.TrimSpace(lower[abs+len(stem):])
+			if !hasAnyPrefix(tail, successNegationTails) {
+				return selfReportReason(strings.TrimSpace(stem) + " …")
 			}
+			start = abs + len(stem)
 		}
 	}
 	return ""
 }
 
-func containsAny(s string, subs []string) bool {
-	for _, sub := range subs {
-		if strings.Contains(s, sub) {
-			return true
-		}
-	}
-	return false
-}
-
 func selfReportReason(marker string) string {
 	return `the final message admits the objective was not met ("` + marker + `")`
 }
@@ -305,8 +230,8 @@ func hasAnyPrefix(s string, prefixes []string) bool {
 // the bug-hunt surfaced: well-formed output treated as correct; pre-existing tests
 // passing treated as the objective being met; and a result that merely matches a
 // baseline the task asked to beat or improve. General — no task-specific content.
-func acceptanceVerificationNudge(objective string) string {
-	nudge := "Before this task can be marked complete, " + acceptanceNudgeMarker + " — " +
+func acceptanceVerificationNudge() string {
+	return "Before this task can be marked complete, " + acceptanceNudgeMarker + " — " +
 		"NOT the shape or format of your output, NOT that pre-existing tests pass, and NOT that your " +
 		"result merely matches a baseline you were asked to beat or improve. " +
 		"Re-read the original task, then run a concrete check that exercises the actual requirement: " +
@@ -314,10 +239,6 @@ func acceptanceVerificationNudge(objective string) string {
 		"thing the task asked you to produce, recover, fix, or optimize. " +
 		"If that check passes, reply PASS and cite the evidence. " +
 		"If it does not pass — or you cannot run such a check — say so plainly and keep working; do not claim success."
-	if objective = capTaskObjective(objective); objective != "" {
-		nudge += "\n\nTask objective: " + objective
-	}
-	return nudge
 }
 
 // toolFailureHintMarker is a stable substring for tests.
@@ -555,12 +476,16 @@ func (state *guardState) pendingPlanItems() bool {
 // many items are still remaining. Malformed arguments leave the prior count
 // unchanged (best-effort — the plan panel itself tolerates the same).
 func (state *guardState) observePlanUpdate(arguments string) {
-	plan, ok := parseTaskPlan(arguments)
-	if !ok {
+	var parsed struct {
+		Plan []struct {
+			Status string `json:"status"`
+		} `json:"plan"`
+	}
+	if json.Unmarshal([]byte(arguments), &parsed) != nil {
 		return
 	}
 	pending := 0
-	for _, item := range plan {
+	for _, item := range parsed.Plan {
 		if planStatusRemaining(item.Status) {
 			pending++
 		}

internal/agent/guardrails_test.go

@@ -138,7 +138,7 @@ func TestRunResetsEmptyTurnCounterOnToolCall(t *testing.T) {
 	root := t.TempDir()
 	writeAgentTestFile(t, root+"/notes.txt", "alpha")
 	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedReadFileTool(root, nil))
+	registry.Register(tools.NewReadFileTool(root))
 
 	provider := &mockProvider{
 		turns: [][]zeroruntime.StreamEvent{
@@ -169,37 +169,6 @@ func TestRunResetsEmptyTurnCounterOnToolCall(t *testing.T) {
 	}
 }
 
-// The whole point of #702's fix: the unknown-session error's normalized
-// signature must not vary with the session id, so a model probing ids 1, 2,
-// 3, … produces one repeated signature the failure guard can count. If the id
-// leaked into the first 80 normalized chars, each probe would reset the streak
-// and the halt would never fire.
-func TestUnknownExecSessionErrorSignatureIsIDInvariant(t *testing.T) {
-	a := errorSignature(tools.UnknownExecSessionError(1))
-	b := errorSignature(tools.UnknownExecSessionError(999999))
-	if a != b {
-		t.Fatalf("unknown-session signature varies with id:\n  %q\n  %q", a, b)
-	}
-}
-
-// End to end: with an id-invariant signature, probing a different unknown id
-// each turn now trips the repeated-failure halt at toolFailureStopAt, where
-// before the fix it never would.
-func TestUnknownExecSessionProbingTripsFailureHalt(t *testing.T) {
-	var state guardState
-	var stoppedAt int
-	for i := 1; i <= toolFailureStopAt; i++ {
-		out := state.observeToolResult(tools.WriteStdinToolName, true, tools.UnknownExecSessionError(i))
-		if out.Stop {
-			stoppedAt = i
-			break
-		}
-	}
-	if stoppedAt != toolFailureStopAt {
-		t.Fatalf("probing distinct unknown ids stopped at %d, want %d", stoppedAt, toolFailureStopAt)
-	}
-}
-
 func TestGuardStateResetsToolOnlyStreakOnEmptyNonToolTurn(t *testing.T) {
 	var state guardState
 	toolOnly := zeroruntime.CollectedStream{
@@ -260,7 +229,7 @@ func TestRunInjectsPlanNotCalledReminderForMultiStepTask(t *testing.T) {
 	root := t.TempDir()
 	writeAgentTestFile(t, root+"/notes.txt", "alpha")
 	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedReadFileTool(root, nil))
+	registry.Register(tools.NewReadFileTool(root))
 
 	provider := &mockProvider{
 		turns: [][]zeroruntime.StreamEvent{
@@ -291,7 +260,7 @@ func TestRunDoesNotInjectPlanReminderForTrivialTask(t *testing.T) {
 	root := t.TempDir()
 	writeAgentTestFile(t, root+"/notes.txt", "alpha")
 	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedReadFileTool(root, nil))
+	registry.Register(tools.NewReadFileTool(root))
 
 	provider := &mockProvider{
 		turns: [][]zeroruntime.StreamEvent{
@@ -319,7 +288,7 @@ func TestRunDoesNotInjectNotCalledReminderWhenPlanUsed(t *testing.T) {
 	root := t.TempDir()
 	writeAgentTestFile(t, root+"/notes.txt", "alpha")
 	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedReadFileTool(root, nil))
+	registry.Register(tools.NewReadFileTool(root))
 	registry.Register(tools.NewUpdatePlanTool())
 
 	provider := &mockProvider{
@@ -349,7 +318,7 @@ func TestRunInjectsStalePlanReminderAfterManyToolCalls(t *testing.T) {
 	root := t.TempDir()
 	writeAgentTestFile(t, root+"/notes.txt", "alpha")
 	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedReadFileTool(root, nil))
+	registry.Register(tools.NewReadFileTool(root))
 	registry.Register(tools.NewUpdatePlanTool())
 
 	// Turn 1 calls update_plan (so the not-called reminder never triggers), then
@@ -383,7 +352,7 @@ func TestRunStalePlanReminderIsOneShotPerInterval(t *testing.T) {
 	root := t.TempDir()
 	writeAgentTestFile(t, root+"/notes.txt", "alpha")
 	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedReadFileTool(root, nil))
+	registry.Register(tools.NewReadFileTool(root))
 	registry.Register(tools.NewUpdatePlanTool())
 
 	turns := [][]zeroruntime.StreamEvent{
@@ -418,7 +387,7 @@ func TestRunInjectsToolOnlyProgressReminder(t *testing.T) {
 	root := t.TempDir()
 	writeAgentTestFile(t, root+"/notes.txt", "alpha")
 	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedReadFileTool(root, nil))
+	registry.Register(tools.NewReadFileTool(root))
 
 	turns := make([][]zeroruntime.StreamEvent, 0, toolOnlyProgressReminderAt+1)
 	for i := 0; i < toolOnlyProgressReminderAt; i++ {

internal/agent/output_budget_propagation_test.go

@@ -1,255 +0,0 @@
-package agent
-
-import (
-	"context"
-	"os"
-	"path/filepath"
-	"strconv"
-	"strings"
-	"testing"
-
-	"github.com/Gitlawb/zero/internal/hooks"
-	"github.com/Gitlawb/zero/internal/tools"
-	"github.com/Gitlawb/zero/internal/trace"
-	"github.com/Gitlawb/zero/internal/zeroruntime"
-)
-
-type propagationOutputTool struct {
-	output string
-}
-
-func TestOutputBudgetHookHelperProcess(t *testing.T) {
-	for index, arg := range os.Args {
-		if arg == "--zero-output-budget-hook-repeat" && index+1 < len(os.Args) {
-			repeats, err := strconv.Atoi(os.Args[index+1])
-			if err != nil {
-				os.Exit(2)
-			}
-			if _, err := os.Stdout.WriteString(strings.Repeat("hook feedback ", repeats)); err != nil {
-				os.Exit(2)
-			}
-			os.Exit(0)
-		}
-		if arg != "--zero-output-budget-hook" || index+1 >= len(os.Args) {
-			continue
-		}
-		if _, err := os.Stdout.WriteString(os.Args[index+1]); err != nil {
-			os.Exit(2)
-		}
-		os.Exit(0)
-	}
-}
-
-func largeOutputBudgetHookDispatcher() *hooks.Dispatcher {
-	return outputBudgetHookDispatcherFor("propagation_output", 200)
-}
-
-func outputBudgetHookDispatcherFor(toolName string, repeats int) *hooks.Dispatcher {
-	feedback := strings.Repeat("hook feedback ", repeats)
-	hookArgs := []string{
-		"-test.run=TestOutputBudgetHookHelperProcess",
-		"--",
-		"--zero-output-budget-hook",
-		feedback,
-	}
-	if repeats > 1_000 {
-		hookArgs = []string{
-			"-test.run=TestOutputBudgetHookHelperProcess",
-			"--",
-			"--zero-output-budget-hook-repeat",
-			strconv.Itoa(repeats),
-		}
-	}
-	return hooks.NewDispatcher(hooks.DispatcherOptions{Config: hooks.Config{
-		Enabled: true,
-		Hooks: []hooks.Definition{{
-			ID:      "large-feedback",
-			Event:   hooks.EventAfterTool,
-			Matcher: toolName,
-			Command: os.Args[0],
-			Args:    hookArgs,
-			Enabled: true,
-		}},
-	}})
-}
-
-func TestExecuteToolCallCommitsReadObservationAfterFinalBoundary(t *testing.T) {
-	root := t.TempDir()
-	targetPath := filepath.Join(root, "target.txt")
-	if err := os.WriteFile(targetPath, []byte("known content\n"), 0o600); err != nil {
-		t.Fatal(err)
-	}
-	linkPath := filepath.Join(root, "small.txt")
-	if err := os.Symlink(targetPath, linkPath); err != nil {
-		t.Skipf("symlink unavailable: %v", err)
-	}
-	resolvedPath, err := filepath.EvalSymlinks(linkPath)
-	if err != nil {
-		t.Fatal(err)
-	}
-	registry := tools.NewRegistry()
-	registry.Register(tools.NewReadFileTool(root))
-	tracker := tools.NewFileTracker()
-
-	result, abortErr := executeToolCall(context.Background(), registry, ToolCall{
-		ID: "read-small", Name: "read_file", Arguments: `{"path":"small.txt"}`,
-	}, PermissionModeAuto, Options{Cwd: root, FileTracker: tracker})
-	if abortErr != nil || result.Status != tools.StatusOK {
-		t.Fatalf("executeToolCall: result=%#v err=%v", result, abortErr)
-	}
-	if !tracker.SeenWhole(resolvedPath) {
-		t.Fatal("unmodified final read output was not committed as observed")
-	}
-}
-
-func TestExecuteToolCallDoesNotCommitReadObservationBeforeHookBudget(t *testing.T) {
-	t.Setenv("ZERO_TOOL_OUTPUT_CEILING_TOKENS", "80")
-	root := t.TempDir()
-	targetPath := filepath.Join(root, "target.txt")
-	if err := os.WriteFile(targetPath, []byte("known content\n"), 0o600); err != nil {
-		t.Fatal(err)
-	}
-	linkPath := filepath.Join(root, "small.txt")
-	if err := os.Symlink(targetPath, linkPath); err != nil {
-		t.Skipf("symlink unavailable: %v", err)
-	}
-	resolvedPath, err := filepath.EvalSymlinks(linkPath)
-	if err != nil {
-		t.Fatal(err)
-	}
-	registry := tools.NewRegistry()
-	registry.Register(tools.NewReadFileTool(root))
-	tracker := tools.NewFileTracker()
-
-	result, abortErr := executeToolCall(context.Background(), registry, ToolCall{
-		ID: "read-hook", Name: "read_file", Arguments: `{"path":"small.txt"}`,
-	}, PermissionModeAuto, Options{
-		Cwd: root, FileTracker: tracker, Hooks: outputBudgetHookDispatcherFor("read_file", 20_000),
-	})
-	if abortErr != nil {
-		t.Fatalf("executeToolCall: %v", abortErr)
-	}
-	if !result.Truncated {
-		t.Fatalf("precondition: hook feedback was not truncated: %#v", result)
-	}
-	if tracker.SeenWhole(resolvedPath) {
-		t.Fatal("read observation committed before hook output was budgeted")
-	}
-}
-
-func (tool propagationOutputTool) Name() string        { return "propagation_output" }
-func (tool propagationOutputTool) Description() string { return "returns output for propagation tests" }
-func (tool propagationOutputTool) Parameters() tools.Schema {
-	return tools.Schema{Type: "object", AdditionalProperties: false}
-}
-func (tool propagationOutputTool) Safety() tools.Safety {
-	return tools.Safety{SideEffect: tools.SideEffectRead, Permission: tools.PermissionAllow, Reason: "test read"}
-}
-func (tool propagationOutputTool) Run(context.Context, map[string]any) tools.Result {
-	return tools.Result{Status: tools.StatusOK, Output: tool.output}
-}
-
-func TestExecuteToolCallPropagatesOutputTruncation(t *testing.T) {
-	t.Setenv("TMPDIR", t.TempDir())
-	t.Setenv("ZERO_TOOL_OUTPUT_CEILING_TOKENS", "80")
-	registry := tools.NewRegistry()
-	registry.Register(propagationOutputTool{output: strings.Repeat("large output\n", 1000)})
-
-	result, abortErr := executeToolCall(context.Background(), registry, ToolCall{
-		ID:        "call-budget",
-		Name:      "propagation_output",
-		Arguments: `{}`,
-	}, PermissionModeAuto, Options{Cwd: t.TempDir()})
-	if abortErr != nil {
-		t.Fatalf("executeToolCall abort error: %v", abortErr)
-	}
-	if !result.Truncated {
-		t.Fatalf("agent ToolResult lost tools.Result.Truncated: %#v", result)
-	}
-	if result.Meta["spill_path"] == "" {
-		t.Fatalf("agent ToolResult lost spill metadata: %#v", result.Meta)
-	}
-}
-
-func TestRecordOutputBudgetTraceUsesOnlyCompactMetadata(t *testing.T) {
-	recorder := trace.NewRecorder("session", "run", "")
-	recorder.Start()
-	recordOutputBudgetTrace(recorder, ToolResult{
-		Name:      "grep",
-		Truncated: true,
-		Output:    "SECRET OUTPUT MUST NOT ENTER TRACE",
-		Meta: map[string]string{
-			"output_budget_category":                  "search",
-			"output_budget_original_bytes":            "1000",
-			"output_budget_retained_bytes":            "100",
-			"output_budget_estimated_original_tokens": "250",
-			"output_budget_estimated_retained_tokens": "25",
-			"output_budget_reason":                    "semantic_search_budget",
-			"output_budget_spill_created":             "true",
-			"spill_path":                              "/secret/path/not-for-trace",
-		},
-	})
-	events := recorder.Finish().OutputBudgets
-	if len(events) != 1 {
-		t.Fatalf("events = %#v", events)
-	}
-	event := events[0]
-	if event.Tool != "grep" || event.Category != "search" || event.OriginalBytes != 1000 || event.RetainedBytes != 100 || !event.SpillCreated {
-		t.Fatalf("unexpected trace event: %#v", event)
-	}
-}
-
-func TestExecuteToolCallRebudgetsOversizedAfterToolFeedback(t *testing.T) {
-	t.Setenv("ZERO_TOOL_OUTPUT_CEILING_TOKENS", "80")
-	registry := tools.NewRegistry()
-	registry.Register(propagationOutputTool{output: "tool output"})
-	dispatcher := largeOutputBudgetHookDispatcher()
-
-	result, abortErr := executeToolCall(context.Background(), registry, ToolCall{
-		ID:        "call-hook-budget",
-		Name:      "propagation_output",
-		Arguments: `{}`,
-	}, PermissionModeAuto, Options{Hooks: dispatcher})
-	if abortErr != nil {
-		t.Fatalf("executeToolCall abort error: %v", abortErr)
-	}
-	if !result.Truncated || len(result.Output) > 80*4 {
-		t.Fatalf("afterTool feedback bypassed output budget: truncated=%t bytes=%d meta=%#v", result.Truncated, len(result.Output), result.Meta)
-	}
-	if result.Meta["output_budget_category"] == "" || result.Meta["output_budget_retained_bytes"] != strconv.Itoa(len(result.Output)) {
-		t.Fatalf("post-hook budget metadata does not describe final output: %#v", result.Meta)
-	}
-}
-
-func TestRunTraceReflectsPostHookBudget(t *testing.T) {
-	t.Setenv("ZERO_TOOL_OUTPUT_CEILING_TOKENS", "80")
-	registry := tools.NewRegistry()
-	registry.Register(propagationOutputTool{output: "tool output"})
-	dispatcher := largeOutputBudgetHookDispatcher()
-	provider := &mockProvider{turns: [][]zeroruntime.StreamEvent{
-		{
-			{Type: zeroruntime.StreamEventToolCallStart, ToolCallID: "call-hook-trace", ToolName: "propagation_output"},
-			{Type: zeroruntime.StreamEventToolCallDelta, ToolCallID: "call-hook-trace", ArgumentsFragment: `{}`},
-			{Type: zeroruntime.StreamEventToolCallEnd, ToolCallID: "call-hook-trace"},
-			{Type: zeroruntime.StreamEventDone},
-		},
-		{{Type: zeroruntime.StreamEventText, Content: "done"}, {Type: zeroruntime.StreamEventDone}},
-	}}
-	recorder := trace.NewRecorder("session", "run", "")
-	var toolResults []ToolResult
-	if _, err := Run(context.Background(), "budget hook", provider, Options{
-		Registry:     registry,
-		Hooks:        dispatcher,
-		Trace:        recorder,
-		OnToolResult: func(result ToolResult) { toolResults = append(toolResults, result) },
-	}); err != nil {
-		t.Fatalf("Run: %v", err)
-	}
-	if len(toolResults) != 1 || !toolResults[0].Truncated {
-		t.Fatalf("tool result = %#v, want one truncated post-hook result", toolResults)
-	}
-	events := recorder.Finish().OutputBudgets
-	if len(events) != 1 || !events[0].Truncated || events[0].RetainedBytes != len(toolResults[0].Output) {
-		t.Fatalf("trace does not describe final post-hook output: events=%#v result=%#v", events, toolResults[0])
-	}
-}

internal/agent/parallel_tools.go

@@ -10,23 +10,12 @@ import (
 // Parallel read-ahead for tool batches. When a turn requests several
 // independent lookups (read_file + grep + glob is the common shape), executing
 // them one after another serializes pure I/O waits. A consecutive run of
-// capability-safe read-only calls is executed concurrently instead; results are
+// auto-allowed read-only calls is executed concurrently instead; results are
 // then consumed in the original call order, so guard counters, message
 // ordering, abort semantics, and the surface's call/result event pairing are
 // byte-identical to sequential execution. Runs never span a mutating call: a
 // read that follows a write must observe the write, so eligibility is decided
 // per consecutive run, not per batch.
-//
-// Eligibility uses the PR5 tool-effect contract (tools.CapabilitiesOf):
-//
-//	Effect == ReadOnly
-//	AND ThreadSafe == true
-//	AND auto-allowed (no interactive permission prompt on the hot path)
-//	AND no resource-key conflict with an earlier call in the same concurrent window
-//
-// Unknown, mutators, interactive tools, and non-thread-safe reads stay
-// sequential. Empty resource keys do not conflict (ThreadSafe is the safety
-// gate); shared non-empty keys force a batch boundary.
 
 // maxParallelReadTools bounds concurrent read-only tool executions in a turn.
 const maxParallelReadTools = 8
@@ -39,117 +28,28 @@ type precomputedToolResult struct {
 }
 
 // parallelSafeToolCall reports whether call may run concurrently with its
-// neighbors under the PR5 capability contract. Loop-intercepted tools
-// (ask_user, request_permissions) and tool_search (mutates the deferred-tool
-// set) always stay sequential.
+// neighbors: the tool must exist, be side-effect-free (SideEffectRead), and be
+// auto-allowed for these args, so no interactive prompt or workspace mutation
+// is on the hot path. Loop-intercepted tools (ask_user, request_permissions)
+// and tool_search (mutates the deferred-tool set) stay sequential.
 func parallelSafeToolCall(registry *tools.Registry, call ToolCall, options Options) bool {
 	switch call.Name {
 	case "ask_user", tools.RequestPermissionsToolName, tools.ToolSearchToolName:
 		return false
 	}
 	tool, found := registry.Get(call.Name)
-	if !found {
+	if !found || tool.Safety().SideEffect != tools.SideEffectRead {
 		return false
 	}
-	caps := tools.CapabilitiesOf(tool)
-	// Fail-closed: only audited concurrent-safe pure reads.
-	if caps.Effect != tools.EffectReadOnly || !caps.ThreadSafe {
-		return false
-	}
-	args, ok := decodeCallArgs(call)
-	if !ok {
-		return false
+	args := map[string]any{}
+	if call.Arguments != "" {
+		if err := decodeToolArguments(call.Arguments, &args); err != nil {
+			return false
+		}
 	}
 	return effectivePermission(tool, args) == tools.PermissionAllow
 }
 
-// decodeCallArgs decodes tool call JSON arguments. Returns false on malformed
-// input so the call stays sequential (never panics into the parallel path).
-func decodeCallArgs(call ToolCall) (map[string]any, bool) {
-	args := map[string]any{}
-	if call.Arguments == "" {
-		return args, true
-	}
-	if err := decodeToolArguments(call.Arguments, &args); err != nil {
-		return nil, false
-	}
-	return args, true
-}
-
-// resourceKeysForCall returns the conflict keys for a call, or nil when the
-// tool has no ResourceKeys function / no keys for these args.
-func resourceKeysForCall(registry *tools.Registry, call ToolCall) []string {
-	tool, found := registry.Get(call.Name)
-	if !found {
-		return nil
-	}
-	caps := tools.CapabilitiesOf(tool)
-	if caps.ResourceKeys == nil {
-		return nil
-	}
-	args, ok := decodeCallArgs(call)
-	if !ok {
-		return nil
-	}
-	return caps.ResourceKeys(args)
-}
-
-// resourceKeysConflict reports whether two key sets share any non-empty key.
-// Empty key sets never conflict: ThreadSafe is the eligibility gate; keys only
-// refine conflict detection among tools that declare specific resources.
-func resourceKeysConflict(a, b []string) bool {
-	if len(a) == 0 || len(b) == 0 {
-		return false
-	}
-	seen := make(map[string]struct{}, len(a))
-	for _, key := range a {
-		if key == "" {
-			continue
-		}
-		seen[key] = struct{}{}
-	}
-	for _, key := range b {
-		if key == "" {
-			continue
-		}
-		if _, ok := seen[key]; ok {
-			return true
-		}
-	}
-	return false
-}
-
-// extendParallelRun returns the exclusive end index of a consecutive parallel-
-// safe run starting at start. The run stops before the first call that is not
-// parallel-safe or that conflicts on resource keys with any earlier call in
-// the same window (so two read_file calls on the same path stay sequential).
-func extendParallelRun(registry *tools.Registry, calls []ToolCall, start int, options Options) int {
-	if start >= len(calls) || !parallelSafeToolCall(registry, calls[start], options) {
-		return start
-	}
-	end := start + 1
-	keysWindow := [][]string{resourceKeysForCall(registry, calls[start])}
-	for end < len(calls) {
-		if !parallelSafeToolCall(registry, calls[end], options) {
-			break
-		}
-		nextKeys := resourceKeysForCall(registry, calls[end])
-		conflict := false
-		for _, prev := range keysWindow {
-			if resourceKeysConflict(prev, nextKeys) {
-				conflict = true
-				break
-			}
-		}
-		if conflict {
-			break
-		}
-		keysWindow = append(keysWindow, nextKeys)
-		end++
-	}
-	return end
-}
-
 // executeParallelReadBatch runs calls[start:end] concurrently (bounded by
 // maxParallelReadTools) and returns results indexed relative to start. All
 // execution-side callbacks that can fire inside executeToolCall are serialized

internal/agent/parallel_tools_test.go

@@ -57,20 +57,6 @@ func (tool *probeTool) Parameters() tools.Schema {
 func (tool *probeTool) Safety() tools.Safety {
 	return tools.Safety{SideEffect: tool.sideEffect, Permission: tools.PermissionAllow, Reason: "test"}
 }
-
-// Capabilities maps the probe's SideEffect into the PR5 capability contract so
-// parallelSafeToolCall (CapabilitiesOf) can classify reads vs mutators. Without
-// this, probes default to EffectUnknown and never enter the concurrent path.
-func (tool *probeTool) Capabilities() tools.ToolCapabilities {
-	switch tool.sideEffect {
-	case tools.SideEffectRead:
-		return tools.ToolCapabilities{Effect: tools.EffectReadOnly, ThreadSafe: true}
-	case tools.SideEffectWrite:
-		return tools.ToolCapabilities{Effect: tools.EffectWorkspaceWrite, ThreadSafe: false}
-	default:
-		return tools.UnknownCapabilities()
-	}
-}
 func (tool *probeTool) Run(_ context.Context, args map[string]any) tools.Result {
 	id, _ := args["id"].(string)
 	tool.mu.Lock()
@@ -106,11 +92,6 @@ func TestParallelSafeToolCall(t *testing.T) {
 	registry := tools.NewRegistry()
 	registry.Register(&probeTool{name: "probe_read", sideEffect: tools.SideEffectRead})
 	registry.Register(&probeTool{name: "probe_write", sideEffect: tools.SideEffectWrite})
-	// Read-only but not ThreadSafe: must stay sequential under the PR5 gate.
-	registry.Register(&keyedProbeTool{
-		probeTool:  probeTool{name: "probe_read_serial", sideEffect: tools.SideEffectRead},
-		threadSafe: false,
-	})
 
 	call := func(name, args string) ToolCall { return ToolCall{ID: "c", Name: name, Arguments: args} }
 	if !parallelSafeToolCall(registry, call("probe_read", `{"id":"a"}`), Options{}) {
@@ -128,91 +109,6 @@ func TestParallelSafeToolCall(t *testing.T) {
 	if parallelSafeToolCall(registry, call("ask_user", `{}`), Options{}) {
 		t.Fatal("loop-intercepted tools must stay sequential")
 	}
-	if parallelSafeToolCall(registry, call("probe_read_serial", `{"id":"a"}`), Options{}) {
-		t.Fatal("ReadOnly without ThreadSafe must not be parallel-safe")
-	}
-}
-
-func TestResourceKeysConflict(t *testing.T) {
-	if resourceKeysConflict(nil, []string{"file:a"}) {
-		t.Fatal("empty vs non-empty must not conflict")
-	}
-	if resourceKeysConflict([]string{"file:a"}, nil) {
-		t.Fatal("non-empty vs empty must not conflict")
-	}
-	if resourceKeysConflict([]string{"file:a"}, []string{"file:b"}) {
-		t.Fatal("distinct keys must not conflict")
-	}
-	if !resourceKeysConflict([]string{"file:a", "file:b"}, []string{"file:b"}) {
-		t.Fatal("shared key must conflict")
-	}
-	if resourceKeysConflict([]string{"", "file:a"}, []string{""}) {
-		t.Fatal("empty-string keys must be ignored")
-	}
-}
-
-func TestExtendParallelRunResourceKeyBoundary(t *testing.T) {
-	// Two keyed reads on the same path must not share a concurrent window;
-	// distinct paths may batch together.
-	registry := tools.NewRegistry()
-	registry.Register(&keyedProbeTool{
-		probeTool:  probeTool{name: "keyed_read", sideEffect: tools.SideEffectRead},
-		threadSafe: true,
-		keys: func(args map[string]any) []string {
-			id, _ := args["id"].(string)
-			if id == "" {
-				return nil
-			}
-			return []string{"file:" + id}
-		},
-	})
-	calls := []ToolCall{
-		{ID: "1", Name: "keyed_read", Arguments: `{"id":"a"}`},
-		{ID: "2", Name: "keyed_read", Arguments: `{"id":"b"}`},
-		{ID: "3", Name: "keyed_read", Arguments: `{"id":"a"}`}, // conflicts with call 0
-	}
-	// start=0: a and b share no key → run covers [0,2)
-	if end := extendParallelRun(registry, calls, 0, Options{}); end != 2 {
-		t.Fatalf("extend from 0 = %d, want 2 (stop before second file:a)", end)
-	}
-	// start=2: single remaining call
-	if end := extendParallelRun(registry, calls, 2, Options{}); end != 3 {
-		t.Fatalf("extend from 2 = %d, want 3", end)
-	}
-	// Empty keys never conflict: three keyless safe reads form one window.
-	registry.Register(&probeTool{name: "keyless_read", sideEffect: tools.SideEffectRead})
-	keyless := []ToolCall{
-		{ID: "1", Name: "keyless_read", Arguments: `{"id":"x"}`},
-		{ID: "2", Name: "keyless_read", Arguments: `{"id":"y"}`},
-		{ID: "3", Name: "keyless_read", Arguments: `{"id":"z"}`},
-	}
-	if end := extendParallelRun(registry, keyless, 0, Options{}); end != 3 {
-		t.Fatalf("keyless extend = %d, want 3", end)
-	}
-}
-
-// keyedProbeTool is a probe with explicit ResourceKeys / ThreadSafe for
-// planner unit tests.
-type keyedProbeTool struct {
-	probeTool
-	threadSafe bool
-	keys       func(args map[string]any) []string
-}
-
-func (tool *keyedProbeTool) Capabilities() tools.ToolCapabilities {
-	caps := tools.ToolCapabilities{
-		Effect:     tools.EffectReadOnly,
-		ThreadSafe: tool.threadSafe,
-	}
-	if tool.keys != nil {
-		caps.ResourceKeys = tool.keys
-	}
-	// Honor mutator side effects from the embedded probe when used as write.
-	if tool.sideEffect == tools.SideEffectWrite {
-		caps.Effect = tools.EffectWorkspaceWrite
-		caps.ThreadSafe = false
-	}
-	return caps
 }
 
 func TestRunExecutesConsecutiveReadsConcurrently(t *testing.T) {

internal/agent/partition_cache_stable_test.go

@@ -14,7 +14,7 @@ func TestPartitionToolsActiveAppendsLoadedToolAfterEagerBlock(t *testing.T) {
 	// "read_file".
 	root := t.TempDir()
 	registry := tools.NewRegistry()
-	registry.Register(tools.NewScopedReadFileTool(root, nil)) // non-deferred, "read_file"
+	registry.Register(tools.NewReadFileTool(root)) // non-deferred, "read_file"
 	registry.Register(fakeToolSearchTool{})
 	registry.Register(fakeDeferredTool{name: "mcp__srv__alpha", desc: "alpha"})
 	registry.Register(fakeDeferredTool{name: "mcp__srv__beta", desc: "beta"})

internal/agent/permission_additional_perms_test.go

@@ -68,61 +68,3 @@ func TestBuildPermissionEventKeepsAllowForOrdinaryAllowedCommand(t *testing.T) {
 		t.Fatal("shouldRequestPermission must be false for an ordinary sandbox-allowed command")
 	}
 }
-
-func TestBuildPermissionEventReusesCoveredAdditionalPermissions(t *testing.T) {
-	call := ToolCall{ID: "call_3", Name: "bash", Arguments: `{}`}
-	args := map[string]any{
-		"command":             "npm install",
-		"sandbox_permissions": string(tools.SandboxPermissionsWithAdditionalPermissions),
-		"additional_permissions": map[string]any{
-			"network": map[string]any{"enabled": true},
-		},
-	}
-	decision := &sandbox.Decision{Action: sandbox.ActionAllow, AutoAllowed: true}
-
-	if shouldRequestPermission(promptShellTool{}, args, true, decision) {
-		t.Fatal("an already-covered capability must not trigger another prompt")
-	}
-	event, ok := buildPermissionEvent(call, promptShellTool{}, args, true, PermissionModeAsk, Options{}, decision)
-	if !ok {
-		t.Fatal("covered capability should still produce an auditable allow event")
-	}
-	if event.Action != PermissionActionAllow {
-		t.Fatalf("Action = %q, want allow for an already-covered capability", event.Action)
-	}
-}
-
-func TestBuildPermissionEventOmitsGenericShellReasonFromPrompt(t *testing.T) {
-	call := ToolCall{ID: "call_4", Name: "bash", Arguments: `{}`}
-	args := map[string]any{"command": "make test"}
-	decision := &sandbox.Decision{Action: sandbox.ActionPrompt, Reason: "runs a shell command"}
-
-	event, ok := buildPermissionEvent(call, promptShellTool{}, args, false, PermissionModeAsk, Options{}, decision)
-	if !ok {
-		t.Fatal("generic shell approval must produce an event")
-	}
-	if event.Reason != "" {
-		t.Fatalf("user-facing reason = %q, want omitted generic tool text", event.Reason)
-	}
-	if event.DecisionReason != "runs a shell command" {
-		t.Fatalf("decision reason = %q, want raw audit reason retained", event.DecisionReason)
-	}
-}
-
-func TestBuildPermissionEventPrefersPolicyReasonOverModelJustification(t *testing.T) {
-	call := ToolCall{ID: "call_5", Name: "bash", Arguments: `{}`}
-	args := map[string]any{
-		"command":             "curl https://example.test",
-		"sandbox_permissions": string(tools.SandboxPermissionsRequireEscalated),
-		"justification":       "Run without sandboxing.",
-	}
-	decision := &sandbox.Decision{Action: sandbox.ActionPrompt, Reason: sandbox.ReasonNetworkBlocked}
-
-	event, ok := buildPermissionEvent(call, promptShellTool{}, args, false, PermissionModeAsk, Options{}, decision)
-	if !ok {
-		t.Fatal("policy approval must produce an event")
-	}
-	if event.Reason != sandbox.ReasonNetworkBlocked {
-		t.Fatalf("user-facing reason = %q, want policy reason %q", event.Reason, sandbox.ReasonNetworkBlocked)
-	}
-}

internal/agent/prefix_escalation_disclosure_test.go

@@ -1,118 +0,0 @@
-package agent
-
-import (
-	"path/filepath"
-	"testing"
-
-	"github.com/Gitlawb/zero/internal/sandbox"
-	"github.com/Gitlawb/zero/internal/tools"
-)
-
-// escalatingEngine allows unsandboxed escalation: no denied reads, so
-// UnsandboxedExecutionAllowed is true and a prefix approval would rewrite the
-// call to require_escalated.
-func escalatingEngine(t *testing.T) *sandbox.Engine {
-	t.Helper()
-	policy := sandbox.DefaultPolicy()
-	policy.DenyRead = nil
-	return sandbox.NewEngine(sandbox.EngineOptions{WorkspaceRoot: t.TempDir(), Policy: policy})
-}
-
-// nonEscalatingEngine has a denied read, which is exactly the condition under
-// which escalation is refused, so no disclosure should be made.
-func nonEscalatingEngine(t *testing.T) *sandbox.Engine {
-	t.Helper()
-	root := t.TempDir()
-	policy := sandbox.DefaultPolicy()
-	policy.DenyRead = []string{filepath.Join(root, "secret.txt")}
-	return sandbox.NewEngine(sandbox.EngineOptions{WorkspaceRoot: root, Policy: policy})
-}
-
-// THE POINT OF THE FILE. The prompt's disclosure and the args rewrite must
-// agree on every input, because the defect being fixed is precisely that the
-// two disagreed: the rewrite escalated out of the sandbox while the prompt said
-// only "allow command prefix for session".
-//
-// Asserting agreement rather than asserting each side's expected value is what
-// makes this hold up. If someone later widens or narrows the rewrite's guards
-// and forgets the flag, the two diverge here and this fails, which a pair of
-// independent expected-value tests would not catch.
-func TestPrefixEscalationDisclosureMatchesTheArgsRewrite(t *testing.T) {
-	escalating := Options{Sandbox: escalatingEngine(t)}
-	denied := Options{Sandbox: nonEscalatingEngine(t)}
-
-	cases := []struct {
-		name     string
-		toolName string
-		args     map[string]any
-		options  Options
-	}{
-		{"plain bash command", "bash", map[string]any{"command": "go test ./..."}, escalating},
-		{"plain exec_command", "exec_command", map[string]any{"cmd": "go build ./..."}, escalating},
-		{"not a shell tool", "read_file", map[string]any{"path": "go.mod"}, escalating},
-		{"no sandbox configured", "bash", map[string]any{"command": "go test ./..."}, Options{}},
-		{"escalation refused by denied reads", "bash", map[string]any{"command": "go test ./..."}, denied},
-		{
-			"call already asks for escalation",
-			"bash",
-			map[string]any{"command": "go test ./...", "sandbox_permissions": string(tools.SandboxPermissionsRequireEscalated)},
-			escalating,
-		},
-		{
-			"call already asks for additional permissions",
-			"bash",
-			map[string]any{"command": "go test ./...", "sandbox_permissions": string(tools.SandboxPermissionsWithAdditionalPermissions)},
-			escalating,
-		},
-		// An unrecognized value is NOT an opt-out: the rewrite overwrites it, so
-		// the disclosure has to fire. Included because the obvious reading of
-		// the guards says otherwise, and a case asserting the wrong thing here
-		// would have passed silently.
-		{
-			"unrecognized sandbox_permissions value",
-			"bash",
-			map[string]any{"command": "go test ./...", "sandbox_permissions": "all"},
-			escalating,
-		},
-		{"empty args", "bash", map[string]any{}, escalating},
-		{"nil args", "bash", nil, escalating},
-	}
-
-	for _, testCase := range cases {
-		t.Run(testCase.name, func(t *testing.T) {
-			disclosed := shellPrefixApprovalEscalates(testCase.toolName, testCase.args, testCase.options)
-
-			// What the user's approval would actually be turned into.
-			rewritten := shellExecutionArgsForApproval(testCase.toolName, testCase.args, PermissionDecisionAllowPrefix, testCase.options)
-			escalates := shellCommandRequiresEscalated(rewritten) && !shellCommandRequiresEscalated(testCase.args)
-
-			if disclosed != escalates {
-				t.Fatalf("prompt discloses escalation=%v but approving actually escalates=%v; the prompt and the rewrite disagree, which is the consent gap this guards", disclosed, escalates)
-			}
-		})
-	}
-}
-
-// The two prefix decisions are the only ones that escalate, so a disclosure
-// attached to a request offering neither would be a false alarm. Checked
-// against the same predicate the rewrite consults.
-func TestOnlyPrefixDecisionsEscalate(t *testing.T) {
-	escalating := map[PermissionDecisionAction]bool{
-		PermissionDecisionAllowPrefix:       true,
-		PermissionDecisionAlwaysAllowPrefix: true,
-	}
-	for _, action := range []PermissionDecisionAction{
-		PermissionDecisionAllow,
-		PermissionDecisionAllowStrict,
-		PermissionDecisionAllowForSession,
-		PermissionDecisionAllowPrefix,
-		PermissionDecisionAlwaysAllowPrefix,
-		PermissionDecisionAlwaysAllow,
-		PermissionDecisionDeny,
-		PermissionDecisionCancel,
-	} {
-		if got := shellPrefixApprovalBypassesSandbox(action); got != escalating[action] {
-			t.Errorf("decision %q bypasses sandbox=%v, want %v", action, got, escalating[action])
-		}
-	}
-}

internal/agent/profile_controller.go

@@ -1,118 +0,0 @@
-package agent
-
-import (
-	"github.com/Gitlawb/zero/internal/sandbox"
-)
-
-// profileController observes per-turn run signals and decides at most one
-// posture escalation per run for the armed execution profile. A nil policy (or
-// a policy without Escalate) makes every method a no-op, so runs without a
-// profile are byte-identical to before — the same opt-in convention as Trace
-// and SelfCorrect.
-//
-// The controller only ever CHANGES loop policy knobs (turn ceiling, reasoning
-// effort, completion gate) through the values the loop applies at its act
-// point; it never injects messages, never touches the model or session, and
-// never interacts with the model-initiated escalate_model path.
-type profileController struct {
-	policy *PostureEscalation
-
-	escalated     bool
-	uncertainSeen int
-
-	failureTripped bool
-	riskTripped    bool
-	scTripped      bool
-	uncertainTrip  bool
-}
-
-// newProfileController is nil-safe: a nil ProfilePolicy (the default for every
-// existing caller) or a nil Escalate yields a controller that observes and
-// decides nothing.
-func newProfileController(policy *ProfilePolicy) *profileController {
-	if policy == nil {
-		return &profileController{}
-	}
-	return &profileController{policy: policy.Escalate}
-}
-
-// observeToolOutcome watches two signals from an executed tool call: the
-// repeated-failure guard's streak (reusing the guard's own counting, not a
-// second failure model) and the sandbox risk classification of an executed
-// result.
-func (c *profileController) observeToolOutcome(outcome toolFailureOutcome, result ToolResult) {
-	if c.policy == nil || c.escalated {
-		return
-	}
-	if c.policy.OnToolFailureStreak > 0 && outcome.Count >= c.policy.OnToolFailureStreak {
-		c.failureTripped = true
-	}
-	if threshold := riskRank(c.policy.OnRiskyMutation); threshold > 0 &&
-		result.DenialReason == DenialNone &&
-		riskRank(result.Risk.Level) >= threshold {
-		// The call EXECUTED (was not denied): partial failures count too, since
-		// the mutation ran. An unrecognized threshold ranks 0 and disables the
-		// signal instead of matching everything.
-		c.riskTripped = true
-	}
-}
-
-// observeUncertain counts uncertain completion evaluations (continue nudges and
-// semantic checks). Headless-only by construction: the completion gate never
-// runs interactively.
-func (c *profileController) observeUncertain() {
-	if c.policy == nil || c.escalated || c.policy.OnCompletionUncertain <= 0 {
-		return
-	}
-	c.uncertainSeen++
-	if c.uncertainSeen >= c.policy.OnCompletionUncertain {
-		c.uncertainTrip = true
-	}
-}
-
-// observeSelfCorrect watches the post-edit verification outcome. Any failing
-// outcome (correcting, reported, aborted) is a signal; passed and disabled are
-// not.
-func (c *profileController) observeSelfCorrect(outcome Outcome) {
-	if c.policy == nil || c.escalated || !c.policy.OnSelfCorrectFailure {
-		return
-	}
-	switch outcome {
-	case OutcomeCorrecting, OutcomeReported, OutcomeAborted:
-		c.scTripped = true
-	}
-}
-
-// maybeEscalate reports the escalation target exactly once, the first time any
-// armed trigger has fired. Subsequent calls return false for the rest of the
-// run: escalation is one-shot and never de-escalates, so no cooldown state is
-// needed.
-func (c *profileController) maybeEscalate() (PostureEscalation, bool) {
-	if c.policy == nil || c.escalated {
-		return PostureEscalation{}, false
-	}
-	if !c.failureTripped && !c.riskTripped && !c.scTripped && !c.uncertainTrip {
-		return PostureEscalation{}, false
-	}
-	c.escalated = true
-	return *c.policy, true
-}
-
-// riskRank orders sandbox risk levels for threshold comparison. Unknown levels
-// rank 0: an unrecognized RESULT level can never meet a valid threshold, and an
-// unrecognized THRESHOLD is rejected before comparison (rank 0 disables the
-// signal) so a typo in a profile can never make every result match.
-func riskRank(level sandbox.RiskLevel) int {
-	switch level {
-	case sandbox.RiskLow:
-		return 1
-	case sandbox.RiskMedium:
-		return 2
-	case sandbox.RiskHigh:
-		return 3
-	case sandbox.RiskCritical:
-		return 4
-	default:
-		return 0
-	}
-}

internal/agent/profile_controller_test.go

@@ -1,345 +0,0 @@
-package agent
-
-import (
-	"context"
-	"strings"
-	"testing"
-
-	"github.com/Gitlawb/zero/internal/sandbox"
-	"github.com/Gitlawb/zero/internal/tools"
-	"github.com/Gitlawb/zero/internal/trace"
-	"github.com/Gitlawb/zero/internal/zeroruntime"
-)
-
-func TestProfileControllerNilPolicyIsNoOp(t *testing.T) {
-	for _, controller := range []*profileController{
-		newProfileController(nil),
-		newProfileController(&ProfilePolicy{Name: "balanced"}), // Escalate nil
-	} {
-		controller.observeToolOutcome(toolFailureOutcome{Count: 99}, ToolResult{Status: tools.StatusOK, Risk: sandbox.Risk{Level: sandbox.RiskCritical}})
-		controller.observeUncertain()
-		controller.observeSelfCorrect(OutcomeAborted)
-		if _, fired := controller.maybeEscalate(); fired {
-			t.Fatal("nil policy must never escalate")
-		}
-	}
-}
-
-func TestProfileControllerEscalatesOnFailureStreak(t *testing.T) {
-	controller := newProfileController(&ProfilePolicy{Escalate: &PostureEscalation{OnToolFailureStreak: 2, MaxTurns: 80}})
-	controller.observeToolOutcome(toolFailureOutcome{Count: 1}, ToolResult{Status: tools.StatusError})
-	if _, fired := controller.maybeEscalate(); fired {
-		t.Fatal("streak below threshold must not escalate")
-	}
-	controller.observeToolOutcome(toolFailureOutcome{Count: 2}, ToolResult{Status: tools.StatusError})
-	target, fired := controller.maybeEscalate()
-	if !fired || target.MaxTurns != 80 {
-		t.Fatalf("expected escalation with the policy target, got fired=%t target=%+v", fired, target)
-	}
-}
-
-func TestProfileControllerEscalatesOnRiskyMutation(t *testing.T) {
-	controller := newProfileController(&ProfilePolicy{Escalate: &PostureEscalation{OnRiskyMutation: sandbox.RiskHigh}})
-	// Medium risk on an executed result: below threshold.
-	controller.observeToolOutcome(toolFailureOutcome{}, ToolResult{Status: tools.StatusOK, Risk: sandbox.Risk{Level: sandbox.RiskMedium}})
-	if _, fired := controller.maybeEscalate(); fired {
-		t.Fatal("medium risk must not trip a high threshold")
-	}
-	// Critical risk on a DENIED result: never counts (the mutation never ran).
-	controller.observeToolOutcome(toolFailureOutcome{}, ToolResult{Status: tools.StatusError, DenialReason: DenialSandboxBlock, Risk: sandbox.Risk{Level: sandbox.RiskCritical}})
-	if _, fired := controller.maybeEscalate(); fired {
-		t.Fatal("a denied result must not trip the risky-mutation signal")
-	}
-	// Critical risk on an executed PARTIAL FAILURE (error status, no denial):
-	// the mutation ran, so it counts.
-	controller.observeToolOutcome(toolFailureOutcome{}, ToolResult{Status: tools.StatusError, Risk: sandbox.Risk{Level: sandbox.RiskCritical}})
-	if _, fired := controller.maybeEscalate(); !fired {
-		t.Fatal("an executed partial failure at critical risk must trip a high threshold")
-	}
-}
-
-func TestProfileControllerIgnoresInvalidRiskThreshold(t *testing.T) {
-	controller := newProfileController(&ProfilePolicy{Escalate: &PostureEscalation{OnRiskyMutation: sandbox.RiskLevel("hgih")}})
-	controller.observeToolOutcome(toolFailureOutcome{}, ToolResult{Status: tools.StatusOK, Risk: sandbox.Risk{Level: sandbox.RiskCritical}})
-	if _, fired := controller.maybeEscalate(); fired {
-		t.Fatal("an unrecognized threshold must disable the signal, not match every result")
-	}
-}
-
-func TestProfileControllerEscalatesOnSelfCorrectFailure(t *testing.T) {
-	controller := newProfileController(&ProfilePolicy{Escalate: &PostureEscalation{OnSelfCorrectFailure: true}})
-	controller.observeSelfCorrect(OutcomePassed)
-	controller.observeSelfCorrect(OutcomeDisabled)
-	if _, fired := controller.maybeEscalate(); fired {
-		t.Fatal("passing/disabled outcomes must not escalate")
-	}
-	controller.observeSelfCorrect(OutcomeReported)
-	if _, fired := controller.maybeEscalate(); !fired {
-		t.Fatal("a failing verification outcome must escalate")
-	}
-}
-
-func TestProfileControllerEscalatesOnUncertainCompletion(t *testing.T) {
-	controller := newProfileController(&ProfilePolicy{Escalate: &PostureEscalation{OnCompletionUncertain: 2}})
-	controller.observeUncertain()
-	if _, fired := controller.maybeEscalate(); fired {
-		t.Fatal("first uncertain evaluation must not trip a threshold of 2")
-	}
-	controller.observeUncertain()
-	if _, fired := controller.maybeEscalate(); !fired {
-		t.Fatal("second uncertain evaluation must escalate")
-	}
-}
-
-func TestProfileControllerEscalatesAtMostOnce(t *testing.T) {
-	controller := newProfileController(&ProfilePolicy{Escalate: &PostureEscalation{OnToolFailureStreak: 1}})
-	controller.observeToolOutcome(toolFailureOutcome{Count: 1}, ToolResult{Status: tools.StatusError})
-	if _, fired := controller.maybeEscalate(); !fired {
-		t.Fatal("expected first escalation")
-	}
-	controller.observeToolOutcome(toolFailureOutcome{Count: 5}, ToolResult{Status: tools.StatusOK, Risk: sandbox.Risk{Level: sandbox.RiskCritical}})
-	controller.observeUncertain()
-	if _, fired := controller.maybeEscalate(); fired {
-		t.Fatal("escalation must be one-shot per run")
-	}
-}
-
-// TestPostureEscalationOnUncertainCompletion proves the uncertain-path act
-// point: the completion gate's continue branch escapes the end-of-turn tail,
-// so escalation must apply before the continue. A cue turn under the gate
-// counts as uncertain, escalates, and the very next request carries the target
-// effort.
-func TestPostureEscalationOnUncertainCompletion(t *testing.T) {
-	provider := &mockProvider{turns: [][]zeroruntime.StreamEvent{
-		{
-			{Type: zeroruntime.StreamEventText, Content: "Let me read the file:"},
-			{Type: zeroruntime.StreamEventDone},
-		},
-		{
-			{Type: zeroruntime.StreamEventText, Content: "Done. All set."},
-			{Type: zeroruntime.StreamEventDone},
-		},
-	}}
-
-	recorder := trace.NewRecorder("posture-session", "run-2", "fast")
-	result, err := Run(context.Background(), "go", provider, Options{
-		Registry:                tools.NewRegistry(),
-		MaxTurns:                10,
-		RequireCompletionSignal: true,
-		Trace:                   recorder,
-		Profile: &ProfilePolicy{
-			Name: "fast",
-			Escalate: &PostureEscalation{
-				ReasoningEffort:       "high",
-				OnCompletionUncertain: 1,
-			},
-		},
-	})
-	if err != nil {
-		t.Fatal(err)
-	}
-	if result.FinalAnswer != "Done. All set." {
-		t.Fatalf("expected the completed answer, got %q", result.FinalAnswer)
-	}
-	if len(provider.requests) != 2 {
-		t.Fatalf("expected 2 requests, got %d", len(provider.requests))
-	}
-	if provider.requests[1].ReasoningEffort != "high" {
-		t.Fatalf("expected the post-escalation request to carry the target effort, got %q", provider.requests[1].ReasoningEffort)
-	}
-	tr := recorder.Finish()
-	if got := tr.Counter(trace.CounterPostureEscalations); got != 1 {
-		t.Fatalf("posture_escalations = %d, want 1", got)
-	}
-}
-
-// A profile that filled the run's effort displaced the provider default ("");
-// RestoreDefaultEffort lets escalation restore that default, which a plain ""
-// ReasoningEffort target cannot express (it means "leave untouched").
-func TestPostureEscalationRestoresDefaultEffort(t *testing.T) {
-	provider := &mockProvider{turns: [][]zeroruntime.StreamEvent{
-		{
-			{Type: zeroruntime.StreamEventText, Content: "Let me read the file:"},
-			{Type: zeroruntime.StreamEventDone},
-		},
-		{
-			{Type: zeroruntime.StreamEventText, Content: "Done. All set."},
-			{Type: zeroruntime.StreamEventDone},
-		},
-	}}
-
-	result, err := Run(context.Background(), "go", provider, Options{
-		Registry:                tools.NewRegistry(),
-		MaxTurns:                10,
-		ReasoningEffort:         "low",
-		RequireCompletionSignal: true,
-		Profile: &ProfilePolicy{
-			Name: "fast",
-			Escalate: &PostureEscalation{
-				RestoreDefaultEffort:  true,
-				OnCompletionUncertain: 1,
-			},
-		},
-	})
-	if err != nil {
-		t.Fatal(err)
-	}
-	if result.FinalAnswer != "Done. All set." {
-		t.Fatalf("expected the completed answer, got %q", result.FinalAnswer)
-	}
-	if len(provider.requests) != 2 {
-		t.Fatalf("expected 2 requests, got %d", len(provider.requests))
-	}
-	if provider.requests[0].ReasoningEffort != "low" {
-		t.Fatalf("pre-escalation request effort = %q, want the profile's low", provider.requests[0].ReasoningEffort)
-	}
-	if provider.requests[1].ReasoningEffort != "" {
-		t.Fatalf("post-escalation request effort = %q, want the restored provider default (empty)", provider.requests[1].ReasoningEffort)
-	}
-}
-
-// failingProfileTool always errors with the same signature so the repeated-
-// failure guard builds a streak.
-type failingProfileTool struct{}
-
-func (failingProfileTool) Name() string        { return "flaky_probe" }
-func (failingProfileTool) Description() string { return "always fails identically (test)" }
-func (failingProfileTool) Parameters() tools.Schema {
-	return tools.Schema{Type: "object", AdditionalProperties: false}
-}
-func (failingProfileTool) Safety() tools.Safety {
-	return tools.Safety{SideEffect: tools.SideEffectRead, Permission: tools.PermissionAllow}
-}
-func (failingProfileTool) Run(context.Context, map[string]any) tools.Result {
-	return tools.Result{Status: tools.StatusError, Output: "probe exploded: same signature"}
-}
-
-// TestPostureEscalationRaisesTurnCeilingMidRun proves the end-to-end act path:
-// a Fast-style run with a 2-turn ceiling hits a failure streak, escalates, and
-// finishes on a turn the original ceiling would have cut off, with the effort
-// target applied to the post-escalation request and the counter emitted.
-func TestPostureEscalationRaisesTurnCeilingMidRun(t *testing.T) {
-	registry := tools.NewRegistry()
-	registry.Register(failingProfileTool{})
-
-	toolTurn := []zeroruntime.StreamEvent{
-		{Type: zeroruntime.StreamEventToolCallStart, ToolCallID: "c1", ToolName: "flaky_probe"},
-		{Type: zeroruntime.StreamEventToolCallEnd, ToolCallID: "c1"},
-		{Type: zeroruntime.StreamEventDone},
-	}
-	provider := &mockProvider{turns: [][]zeroruntime.StreamEvent{
-		toolTurn,
-		toolTurn,
-		{
-			{Type: zeroruntime.StreamEventText, Content: "recovered after escalation"},
-			{Type: zeroruntime.StreamEventDone},
-		},
-	}}
-
-	recorder := trace.NewRecorder("posture-session", "run-1", "fast")
-	result, err := Run(context.Background(), "go", provider, Options{
-		Registry: registry,
-		MaxTurns: 2,
-		Trace:    recorder,
-		Profile: &ProfilePolicy{
-			Name: "fast",
-			Escalate: &PostureEscalation{
-				MaxTurns:            4,
-				ReasoningEffort:     "high",
-				OnToolFailureStreak: 2,
-			},
-		},
-	})
-	if err != nil {
-		t.Fatal(err)
-	}
-	if result.FinalAnswer != "recovered after escalation" {
-		t.Fatalf("expected the run to continue past the original 2-turn ceiling, got %q (turns=%d)", result.FinalAnswer, result.Turns)
-	}
-	if len(provider.requests) != 3 {
-		t.Fatalf("expected 3 requests (ceiling raised from 2 to 4), got %d", len(provider.requests))
-	}
-	if provider.requests[2].ReasoningEffort != "high" {
-		t.Fatalf("expected the post-escalation request to carry the target effort, got %q", provider.requests[2].ReasoningEffort)
-	}
-	tr := recorder.Finish()
-	if got := tr.Counter(trace.CounterPostureEscalations); got != 1 {
-		t.Fatalf("posture_escalations = %d, want 1", got)
-	}
-}
-
-// TestPostureEscalationAbsentWithoutProfile pins the no-regression invariant at
-// the loop level: the identical failing script without a profile ends at the
-// original ceiling with the max-turns answer.
-func TestPostureEscalationAbsentWithoutProfile(t *testing.T) {
-	registry := tools.NewRegistry()
-	registry.Register(failingProfileTool{})
-
-	toolTurn := []zeroruntime.StreamEvent{
-		{Type: zeroruntime.StreamEventToolCallStart, ToolCallID: "c1", ToolName: "flaky_probe"},
-		{Type: zeroruntime.StreamEventToolCallEnd, ToolCallID: "c1"},
-		{Type: zeroruntime.StreamEventDone},
-	}
-	provider := &mockProvider{turns: [][]zeroruntime.StreamEvent{toolTurn, toolTurn}}
-
-	result, err := Run(context.Background(), "go", provider, Options{
-		Registry: registry,
-		MaxTurns: 2,
-	})
-	if err != nil {
-		t.Fatal(err)
-	}
-	if len(provider.requests) > 3 {
-		t.Fatalf("unprofiled run must not extend the ceiling, got %d requests", len(provider.requests))
-	}
-	if !strings.Contains(result.FinalAnswer, "maximum number of turns") {
-		t.Fatalf("unprofiled run must end with the max-turns answer, got %q", result.FinalAnswer)
-	}
-}
-
-// TestExecuteToolCallClassifiesRiskWithoutSandbox verifies the executed-risk
-// stamp on the unsandboxed path: a read-only tool classifies low.
-func TestExecuteToolCallClassifiesRiskWithoutSandbox(t *testing.T) {
-	registry := tools.NewRegistry()
-	registry.Register(failingProfileTool{})
-
-	result, abortErr := executeToolCall(
-		context.Background(),
-		registry,
-		ToolCall{ID: "c1", Name: "flaky_probe"},
-		PermissionModeAuto,
-		Options{},
-	)
-	if abortErr != nil {
-		t.Fatal(abortErr)
-	}
-	if result.Risk.Level == "" {
-		t.Fatal("executed result must carry a risk classification")
-	}
-	if riskRank(result.Risk.Level) > riskRank(sandbox.RiskMedium) {
-		t.Fatalf("read-only tool classified %q, want at most medium", result.Risk.Level)
-	}
-}
-
-// TestUnknownToolResultCarriesZeroRisk verifies a call that never executed a
-// tool (unknown name, nil tool) keeps the zero risk value instead of panicking
-// or inventing a classification.
-func TestUnknownToolResultCarriesZeroRisk(t *testing.T) {
-	registry := tools.NewRegistry()
-	result, abortErr := executeToolCall(
-		context.Background(),
-		registry,
-		ToolCall{ID: "c1", Name: "no_such_tool"},
-		PermissionModeAuto,
-		Options{},
-	)
-	if abortErr != nil {
-		t.Fatal(abortErr)
-	}
-	if result.Status == tools.StatusOK {
-		t.Fatal("expected an error result for an unknown tool")
-	}
-	if result.Risk.Level != "" {
-		t.Fatalf("not-executed result must keep the zero risk value, got %q", result.Risk.Level)
-	}
-}

internal/agent/prompt_budget_test.go

@@ -14,26 +14,13 @@ import (
 // deliberately) or trimmed — the per-turn floor should not creep up silently.
 //
 // Measured baselines (2026-07): base system prompt ~3160 tokens; the tools a normal
-// (auto permission-mode) interactive turn sends ~3230 tokens. The tool figure is
-// the auto-mode advertised set, including exec_command: auto mode must expose both
-// process creation and write_stdin process interaction or the latter is unusable
-// without a session id. Higher-risk tools that do not opt into auto remain excluded.
+// (auto permission-mode) interactive turn sends ~2430 tokens. The tool figure is
+// the auto-mode advertised set, not the full core registry — higher-risk tools that
+// only advertise in other modes are excluded, matching what a real turn actually
+// pays.
 const (
 	maxBaseSystemPromptTokens = 3500
-	// Raised deliberately from 3550 on 2026-08-08. view_image (#843) took the set
-	// to 3588, and #867 trimmed it back to 3578: still 28 tokens over.
-	//
-	// The overrun is a rounding error rather than a regression. view_image costs
-	// 82 tokens, the second cheapest tool in the set, against exec_command's 909
-	// and request_permissions' 397; it is simply the one that crossed the line.
-	// Raised rather than paid down because the alternative was trimming a
-	// description to claw back 28 tokens, degrading a tool's usability to satisfy
-	// a line drawn against a 2026-07 measurement that predates two tools.
-	//
-	// The ratchet did its job: it caught the creep and forced a decision instead
-	// of a drift. That is the point of it, so keep raising it deliberately rather
-	// than reflexively.
-	maxEagerToolSchemaTokens = 3650
+	maxEagerToolSchemaTokens  = 2700
 )
 
 func TestSystemPromptTokenBudget(t *testing.T) {
@@ -59,11 +46,6 @@ func TestEagerToolSchemaTokenBudget(t *testing.T) {
 	got := estimateToolDefTokens(exposed)
 	t.Logf("eager core tool schemas: %d tokens across %d tools", got, len(exposed))
 	if got > maxEagerToolSchemaTokens {
-		// NOT "defer a tool": this test pins DeferThreshold at 0, so marking a
-		// tool deferred leaves it exposed here and changes nothing. Deferral is
-		// still worth doing for real sessions, it just cannot move this number.
-		// The levers that do are a smaller schema, one fewer core tool, or a
-		// deliberate raise.
-		t.Fatalf("eager tool schemas are %d tokens, over the %d ceiling — trim a schema, drop a core tool, or raise the ceiling deliberately (deferring will NOT help: this test disables deferral)", got, maxEagerToolSchemaTokens)
+		t.Fatalf("eager tool schemas are %d tokens, over the %d ceiling — defer a tool or raise the ceiling deliberately", got, maxEagerToolSchemaTokens)
 	}
 }

internal/agent/prompt_fingerprint.go

@@ -1,244 +0,0 @@
-package agent
-
-import (
-	"crypto/sha256"
-	"encoding/hex"
-	"encoding/json"
-	"fmt"
-	"sort"
-	"strings"
-
-	"github.com/Gitlawb/zero/internal/zeroruntime"
-)
-
-// promptSubstrings are the cacheable components of the prompt a run sends to a
-// model. systemPrompt is the exact joined system-message content. The remaining
-// fields retain useful diagnostic boundaries within that prompt and tool list.
-type promptSubstrings struct {
-	systemPrompt       string
-	baseInstructions   string
-	confirmationPolicy string
-	projectContext     string
-	skills             string
-	tools              string
-	schema             string
-}
-
-// buildPromptSubstrings builds the prompt and retains both its exact joined
-// form and narrower diagnostic components. Callers that already built the
-// prompt should use buildPromptSubstringsFromParts to avoid repeating workspace
-// reads.
-func buildPromptSubstrings(options Options, exposed []zeroruntime.ToolDefinition) promptSubstrings {
-	return buildPromptSubstringsFromParts(buildSystemPromptParts(options), exposed)
-}
-
-func buildPromptSubstringsFromParts(parts systemPromptParts, exposed []zeroruntime.ToolDefinition) promptSubstrings {
-	toolsSubstr, schemaSubstr := toolSubstrings(exposed)
-
-	return promptSubstrings{
-		systemPrompt:       parts.prompt,
-		baseInstructions:   parts.baseInstructions,
-		confirmationPolicy: parts.confirmationPolicy,
-		projectContext:     parts.projectContext,
-		skills:             parts.skills,
-		tools:              toolsSubstr,
-		schema:             schemaSubstr,
-	}
-}
-
-// toolSubstrings preserves the emitted tool order. Names and descriptions are
-// fingerprinted separately from schemas so either kind of drift is identifiable.
-// Order is part of request identity: silently sorting here would hide a provider-
-// visible reordering and could report a false cache hit.
-func toolSubstrings(exposed []zeroruntime.ToolDefinition) (toolsSubstr, schemaSubstr string) {
-	if len(exposed) == 0 {
-		return "", ""
-	}
-	var toolsSB, schemaSB strings.Builder
-	for _, def := range exposed {
-		writeLengthPrefixed(&toolsSB, def.Name)
-		writeLengthPrefixed(&toolsSB, def.Description)
-		// Parameters is rendered to a canonical JSON string per tool. The
-		// schema-render cache in internal/agent (used by partitionToolsCached)
-		// guarantees the same render is byte-identical across turns for the
-		// same tool, so this substring is a stable hash input.
-		writeLengthPrefixed(&schemaSB, def.Name)
-		writeLengthPrefixed(&schemaSB, canonicalSchemaString(def.Parameters))
-	}
-	return toolsSB.String(), schemaSB.String()
-}
-
-func writeLengthPrefixed(sb *strings.Builder, value string) {
-	fmt.Fprintf(sb, "%d:", len(value))
-	sb.WriteString(value)
-}
-
-// canonicalSchemaString renders a tool's parameter schema to a stable string.
-// tools.ToolDef.Parameters is a map[string]any in the common case; Go's
-// fmt.Sprintf("%v", m) iterates a map in random order, which would produce
-// a different hash for the same Parameters value across calls and defeat
-// the fingerprint. encoding/json marshals maps with keys sorted
-// alphabetically, so json.Marshal is the primary stable render. The
-// fallback for the rare non-JSON-compatible value (functions, channels,
-// NaN/Inf floats, cyclic references) is a stable key-sorted stringifier
-// that walks the value with sorted keys for maps and a leading
-// "__non_json:" prefix so a future schema change is visible in the trace
-// (a SchemaHash collision from the same value) rather than a silent hash
-// drift.
-//
-// Note: SchemaHash is a stability signal, not a wire-identity signal. The
-// bytes json.Marshal produces for a Go map[string]any may differ from
-// the bytes the provider actually sends on the wire (provider encoders
-// use their own JSON conventions; an Anthropic schema may serialize with
-// different whitespace, key order, or number formatting). A consumer
-// must NOT assume SchemaHash matches the provider's wire schema — only
-// that two turns with the same Go-side Parameters produce the same
-// SchemaHash. This is sufficient for the trace's purpose (drift
-// detection) but not for a content-based equality check.
-func canonicalSchemaString(params map[string]any) string {
-	if len(params) == 0 {
-		return ""
-	}
-	data, err := json.Marshal(params)
-	if err != nil {
-		return "__non_json:" + stableStringify(params)
-	}
-	return string(data)
-}
-
-// stableStringify renders v to a deterministic string. Maps have their keys
-// sorted alphabetically; slices are walked in order; primitives use their
-// natural Go format. The function is recursive but bounded by the size of
-// the input, so a cyclic reference is not reachable here (the json.Marshal
-// fallback is hit when json.Marshal itself returns an error, which for the
-// common case is a non-JSON-compatible value, not a cycle — cycles are
-// rare in tool schemas and acceptable to mis-render in the fallback path
-// since the trace's contract is "the hash is stable for the same input,"
-// not "the fallback is lossless"). Used only when json.Marshal fails.
-func stableStringify(v any) string {
-	var sb strings.Builder
-	writeStable(&sb, v)
-	return sb.String()
-}
-
-func writeStable(sb *strings.Builder, v any) {
-	switch x := v.(type) {
-	case nil:
-		sb.WriteString("null")
-	case bool:
-		if x {
-			sb.WriteString("true")
-		} else {
-			sb.WriteString("false")
-		}
-	case string:
-		sb.WriteByte('"')
-		sb.WriteString(x)
-		sb.WriteByte('"')
-	case float64:
-		// Use %g for compact, deterministic float rendering. NaN and Inf are
-		// not JSON-encodable (which is why we are in the fallback path) and
-		// render as "NaN" / "+Inf" / "-Inf" — distinct, stable strings.
-		fmt.Fprintf(sb, "%g", x)
-	case float32:
-		fmt.Fprintf(sb, "%g", float64(x))
-	case int:
-		fmt.Fprintf(sb, "%d", x)
-	case int64:
-		fmt.Fprintf(sb, "%d", x)
-	case int32:
-		fmt.Fprintf(sb, "%d", x)
-	case uint:
-		fmt.Fprintf(sb, "%d", x)
-	case uint64:
-		fmt.Fprintf(sb, "%d", x)
-	case uint32:
-		fmt.Fprintf(sb, "%d", x)
-	case map[string]any:
-		sb.WriteByte('{')
-		keys := make([]string, 0, len(x))
-		for k := range x {
-			keys = append(keys, k)
-		}
-		sort.Strings(keys)
-		for i, k := range keys {
-			if i > 0 {
-				sb.WriteByte(',')
-			}
-			sb.WriteByte('"')
-			sb.WriteString(k)
-			sb.WriteString(`":`)
-			writeStable(sb, x[k])
-		}
-		sb.WriteByte('}')
-	case []any:
-		sb.WriteByte('[')
-		for i, item := range x {
-			if i > 0 {
-				sb.WriteByte(',')
-			}
-			writeStable(sb, item)
-		}
-		sb.WriteByte(']')
-	default:
-		// Last-resort: include the type so distinct values produce distinct
-		// strings even if their default format collides. Without the type
-		// tag, fmt.Sprintf("%v", x) for two different types could produce
-		// the same bytes (rare, but the type prefix makes it impossible).
-		fmt.Fprintf(sb, "<%T:%v>", x, x)
-	}
-}
-
-// ComputePrefixFingerprint builds and fingerprints a prompt for callers that do
-// not already have one. The agent loop uses computePrefixFingerprint with the
-// prompt parts it built once at run start, ensuring tracing observes the exact
-// request content without repeating workspace I/O.
-func ComputePrefixFingerprint(options Options, exposed []zeroruntime.ToolDefinition) prefixFingerprint {
-	return computePrefixFingerprint(buildPromptSubstrings(options, exposed))
-}
-
-func computePrefixFingerprint(subs promptSubstrings) prefixFingerprint {
-	systemPrompt := sha256hex(subs.systemPrompt)
-	base := sha256hex(subs.baseInstructions)
-	policy := sha256hex(subs.confirmationPolicy)
-	project := sha256hex(subs.projectContext)
-	skills := sha256hex(subs.skills)
-	toolsH := sha256hex(subs.tools)
-	schema := sha256hex(subs.schema)
-	complete := sha256hex(strings.Join([]string{
-		systemPrompt, toolsH, schema,
-	}, "|"))
-	return prefixFingerprint{
-		SystemPromptHash:       systemPrompt,
-		BaseInstructionsHash:   base,
-		ConfirmationPolicyHash: policy,
-		ProjectContextHash:     project,
-		SkillsHash:             skills,
-		ToolsHash:              toolsH,
-		SchemaHash:             schema,
-		CompletePrefixHash:     complete,
-	}
-}
-
-// prefixFingerprint is the agent-side shape of a prompt-prefix fingerprint. It
-// is converted to a trace.PrefixHash at the loop boundary (see EmitPrefixHash
-// in loop.go) so the trace package does not need to import this one. The field
-// names match the trace.PrefixHash JSON tags 1:1.
-type prefixFingerprint struct {
-	SystemPromptHash       string
-	BaseInstructionsHash   string
-	ConfirmationPolicyHash string
-	ProjectContextHash     string
-	SkillsHash             string
-	ToolsHash              string
-	SchemaHash             string
-	CompletePrefixHash     string
-}
-
-// sha256hex returns the hex-encoded SHA-256 of s. Empty input produces the
-// hash of the empty string, which is a constant; callers that want to
-// distinguish "absent" from "empty" should check s before calling.
-func sha256hex(s string) string {
-	sum := sha256.Sum256([]byte(s))
-	return hex.EncodeToString(sum[:])
-}

internal/agent/prompt_fingerprint_test.go

@@ -1,264 +0,0 @@
-package agent
-
-import (
-	"strings"
-	"testing"
-
-	"github.com/Gitlawb/zero/internal/zeroruntime"
-)
-
-// TestComputePrefixFingerprintStableAcrossCalls asserts the headline property
-// the trace is meant to expose: two ComputePrefixFingerprint calls over the
-// same Options and the same exposed tool list produce byte-identical hashes.
-// This is the regression catch for "did anyone introduce non-determinism into
-// the prompt prefix."
-func TestComputePrefixFingerprintStableAcrossCalls(t *testing.T) {
-	opts := Options{Cwd: t.TempDir(), SystemPrompt: "core"}
-	exposed := []zeroruntime.ToolDefinition{
-		{Name: "read_file", Parameters: map[string]any{"type": "object"}},
-		{Name: "grep", Parameters: map[string]any{"type": "object"}},
-	}
-	first := ComputePrefixFingerprint(opts, exposed)
-	second := ComputePrefixFingerprint(opts, exposed)
-	if first != second {
-		t.Fatalf("fingerprint must be stable across calls with the same input:\n  first=%+v\n second=%+v", first, second)
-	}
-	if first.CompletePrefixHash == "" {
-		t.Fatalf("CompletePrefixHash must be non-empty for a non-trivial prompt")
-	}
-}
-
-// Tool order and descriptions are provider-visible and therefore part of the
-// cacheable prefix identity.
-func TestComputePrefixFingerprintToolsHashMatchesEmittedOrderAndDescriptions(t *testing.T) {
-	opts := Options{Cwd: t.TempDir(), SystemPrompt: "core"}
-	a := []zeroruntime.ToolDefinition{
-		{Name: "read_file", Description: "read", Parameters: map[string]any{"schema": "schema-read"}},
-		{Name: "grep", Description: "search", Parameters: map[string]any{"schema": "schema-grep"}},
-	}
-	b := []zeroruntime.ToolDefinition{
-		{Name: "grep", Description: "search", Parameters: map[string]any{"schema": "schema-grep"}},
-		{Name: "read_file", Description: "read", Parameters: map[string]any{"schema": "schema-read"}},
-	}
-	fpA := ComputePrefixFingerprint(opts, a)
-	fpB := ComputePrefixFingerprint(opts, b)
-	if fpA.ToolsHash == fpB.ToolsHash {
-		t.Fatalf("ToolsHash must change when emitted order changes: a=%s b=%s", fpA.ToolsHash, fpB.ToolsHash)
-	}
-	c := append([]zeroruntime.ToolDefinition(nil), a...)
-	c[0].Description = "read a file"
-	fpC := ComputePrefixFingerprint(opts, c)
-	if fpA.ToolsHash == fpC.ToolsHash {
-		t.Fatalf("ToolsHash must change when a description changes: a=%s c=%s", fpA.ToolsHash, fpC.ToolsHash)
-	}
-}
-
-// TestComputePrefixFingerprintSchemaChangeChangesSchemaHash asserts that a
-// change to a tool's Parameters schema is visible in the SchemaHash. A schema
-// edit that should be cache-stable (e.g. field reorder, doc tweak) will move
-// the hash and surface in the trace; a schema edit that legitimately
-// invalidates the cache also moves the hash, which is the right behavior.
-func TestComputePrefixFingerprintSchemaChangeChangesSchemaHash(t *testing.T) {
-	opts := Options{Cwd: t.TempDir(), SystemPrompt: "core"}
-	a := []zeroruntime.ToolDefinition{
-		{Name: "read_file", Parameters: map[string]any{"version": "v1"}},
-	}
-	b := []zeroruntime.ToolDefinition{
-		{Name: "read_file", Parameters: map[string]any{"version": "v2"}},
-	}
-	fpA := ComputePrefixFingerprint(opts, a)
-	fpB := ComputePrefixFingerprint(opts, b)
-	if fpA.SchemaHash == fpB.SchemaHash {
-		t.Fatalf("SchemaHash must change when tool Parameters change: a=%s b=%s", fpA.SchemaHash, fpB.SchemaHash)
-	}
-}
-
-// TestComputePrefixFingerprintSystemPromptChangeChangesBaseHash asserts the
-// most common drift path: a system-prompt edit (e.g. an agent.md update, a
-// model addendum change) must move the base-instructions hash, which moves
-// the complete-prefix hash, which is what the trace is for.
-func TestComputePrefixFingerprintSystemPromptChangeChangesBaseHash(t *testing.T) {
-	optsA := Options{Cwd: t.TempDir(), SystemPrompt: "core v1"}
-	optsB := Options{Cwd: t.TempDir(), SystemPrompt: "core v2"}
-	exposed := []zeroruntime.ToolDefinition{{Name: "noop", Parameters: map[string]any{}}}
-	fpA := ComputePrefixFingerprint(optsA, exposed)
-	fpB := ComputePrefixFingerprint(optsB, exposed)
-	if fpA.BaseInstructionsHash == fpB.BaseInstructionsHash {
-		t.Fatalf("BaseInstructionsHash must change when SystemPrompt changes: a=%s b=%s", fpA.BaseInstructionsHash, fpB.BaseInstructionsHash)
-	}
-	if fpA.CompletePrefixHash == fpB.CompletePrefixHash {
-		t.Fatalf("CompletePrefixHash must change when any sub-hash changes: a=%s b=%s", fpA.CompletePrefixHash, fpB.CompletePrefixHash)
-	}
-}
-
-// CompletePrefixHash represents the exact joined system prompt and the ordered
-// tool definitions. Narrower system-section hashes remain diagnostic only.
-func TestComputePrefixFingerprintCompleteIsAggregateOfSubHashes(t *testing.T) {
-	opts := Options{Cwd: t.TempDir(), SystemPrompt: "core"}
-	exposed := []zeroruntime.ToolDefinition{{Name: "read_file", Parameters: map[string]any{"k": "v"}}}
-	fp := ComputePrefixFingerprint(opts, exposed)
-	// Manually compute what the canonical join should be.
-	expected := sha256hex(strings.Join([]string{
-		fp.SystemPromptHash,
-		fp.ToolsHash,
-		fp.SchemaHash,
-	}, "|"))
-	if fp.CompletePrefixHash != expected {
-		t.Fatalf("CompletePrefixHash must equal sha256hex of the exact system, tools, and schema hashes:\n  got:      %s\n  expected: %s", fp.CompletePrefixHash, expected)
-	}
-}
-
-func TestComputePrefixFingerprintCoversEverySystemPromptSection(t *testing.T) {
-	base := Options{Cwd: t.TempDir(), SystemPrompt: "core", Model: "gpt-5"}
-	changed := base
-	changed.ResponseStyle = "concise"
-
-	fpBase := ComputePrefixFingerprint(base, nil)
-	fpChanged := ComputePrefixFingerprint(changed, nil)
-	if fpBase.SystemPromptHash == fpChanged.SystemPromptHash {
-		t.Fatal("SystemPromptHash must change when any rendered system section changes")
-	}
-	if fpBase.CompletePrefixHash == fpChanged.CompletePrefixHash {
-		t.Fatal("CompletePrefixHash must include the exact full system prompt")
-	}
-}
-
-func TestComputePrefixFingerprintCoversTransientSystemPrompt(t *testing.T) {
-	base := Options{Cwd: t.TempDir(), SystemPrompt: "core"}
-	changed := base
-	changed.TransientSystemPrompt = "turn-only guidance"
-
-	fpBase := ComputePrefixFingerprint(base, nil)
-	fpChanged := ComputePrefixFingerprint(changed, nil)
-	if fpBase.SystemPromptHash == fpChanged.SystemPromptHash {
-		t.Fatal("SystemPromptHash must change for transient system guidance")
-	}
-	if fpBase.CompletePrefixHash == fpChanged.CompletePrefixHash {
-		t.Fatal("CompletePrefixHash must cover transient system guidance")
-	}
-}
-
-// TestBuildPromptSubstringsDefaultOptions asserts the invariants of the
-// substrings helper for default Options: which substrings are non-empty,
-// which are empty, and that the substring-to-hash round-trip is lossless.
-func TestBuildPromptSubstringsDefaultOptions(t *testing.T) {
-	opts := Options{
-		Cwd:          t.TempDir(),
-		SystemPrompt: "test core",
-	}
-	subs := buildPromptSubstrings(opts, nil)
-	// Invariants the trace depends on (default Options):
-	//   1. baseInstructions substring equals the core system prompt bytes.
-	//   2. confirmationPolicy substring is non-empty (the embedded policy
-	//      is always present, post TrimSpace).
-	//   3. skills, tools, and schema substrings are empty (no skills or
-	//      tools configured for default Options).
-	if subs.baseInstructions != "test core" {
-		t.Fatalf("baseInstructions substring must equal the core system prompt: got %q want %q", subs.baseInstructions, "test core")
-	}
-	if subs.systemPrompt != buildSystemPrompt(opts) {
-		t.Fatal("systemPrompt substring must equal the exact joined system prompt")
-	}
-	if subs.confirmationPolicy == "" {
-		t.Fatalf("confirmationPolicy substring must be non-empty (the embedded policy is always present)")
-	}
-	if subs.skills != "" {
-		t.Fatalf("skills substring must be empty for default Options: got %q", subs.skills)
-	}
-	// Round-trip: the corresponding fingerprint hashes must equal
-	// sha256hex of the substrings, with no truncation or reformatting
-	// between the two layers.
-	fp := ComputePrefixFingerprint(opts, nil)
-	if fp.SystemPromptHash != sha256hex(subs.systemPrompt) {
-		t.Fatalf("SystemPromptHash in fingerprint must equal sha256hex of the joined prompt")
-	}
-	if fp.BaseInstructionsHash != sha256hex(subs.baseInstructions) {
-		t.Fatalf("BaseInstructionsHash in fingerprint must equal sha256hex of the substring")
-	}
-	if fp.ConfirmationPolicyHash != sha256hex(subs.confirmationPolicy) {
-		t.Fatalf("ConfirmationPolicyHash in fingerprint must equal sha256hex of the substring")
-	}
-}
-
-// TestCanonicalSchemaStringStableForMapParameters asserts the headline
-// property the trace needs: the same map[string]any produces the same
-// canonical string across calls, even though Go's fmt.Sprintf("%v", m)
-// iterates a map in random order. json.Marshal sorts map keys
-// alphabetically, which is the fix.
-func TestCanonicalSchemaStringStableForMapParameters(t *testing.T) {
-	params := map[string]any{
-		"description": "read a file",
-		"properties": map[string]any{
-			"path": map[string]any{"type": "string"},
-		},
-		"required": []any{"path"},
-		"type":     "object",
-	}
-	first := canonicalSchemaString(params)
-	second := canonicalSchemaString(params)
-	if first != second {
-		t.Fatalf("canonicalSchemaString must be stable across calls for the same Parameters:\n  first=%s\n second=%s", first, second)
-	}
-	if first == "" {
-		t.Fatal("canonicalSchemaString must produce a non-empty string for a non-empty map")
-	}
-	// Two maps with the same key/value pairs (different declaration
-	// order in source) must produce identical canonical strings. This
-	// is the property that defeats Go's map iteration randomization.
-	reordered := map[string]any{
-		"type":     "object",
-		"required": []any{"path"},
-		"properties": map[string]any{
-			"path": map[string]any{"type": "string"},
-		},
-		"description": "read a file",
-	}
-	if canonicalSchemaString(params) != canonicalSchemaString(reordered) {
-		t.Fatal("canonicalSchemaString must produce identical output for maps with the same key/value pairs in different declaration orders")
-	}
-}
-
-// TestCanonicalSchemaStringFallbackStableForNonJSONValue exercises the
-// non-JSON fallback in canonicalSchemaString: when Parameters contains a
-// value json.Marshal cannot encode (here, a channel), the function must
-// fall back to a key-sorted stringification that is stable across calls
-// and across map-iteration order. The previous fallback used
-// fmt.Sprintf("%v", params), which iterates the map in random order and
-// produced a different hash for the same value across calls — defeating
-// the fingerprint. The current fallback is stable by construction.
-func TestCanonicalSchemaStringFallbackStableForNonJSONValue(t *testing.T) {
-	ch := make(chan int) // channels are not JSON-encodable
-	params := map[string]any{
-		"description": "a tool with a non-JSON value",
-		"channel":     ch,
-		"z_first":     1,
-		"a_second":    2,
-	}
-	first := canonicalSchemaString(params)
-	second := canonicalSchemaString(params)
-	if first == "" {
-		t.Fatal("canonicalSchemaString must produce a non-empty string for the fallback path")
-	}
-	if first != second {
-		t.Fatalf("canonicalSchemaString fallback must be stable across calls:\n  first=%s\n second=%s", first, second)
-	}
-	// A re-ordered map (same key/value pairs, different declaration order)
-	// must produce the same fallback string. This is the property the
-	// key-sorted stringifier provides that fmt.Sprintf("%v", m) does not.
-	reordered := map[string]any{
-		"a_second":    2,
-		"z_first":     1,
-		"channel":     ch,
-		"description": "a tool with a non-JSON value",
-	}
-	if canonicalSchemaString(params) != canonicalSchemaString(reordered) {
-		t.Fatalf("canonicalSchemaString fallback must produce identical output for maps with the same key/value pairs in different declaration orders:\n  first=%s\n second=%s", first, canonicalSchemaString(reordered))
-	}
-	// The fallback must include the "__non_json:" prefix so a consumer
-	// can tell the bytes are not a JSON-marshaled schema (a hash collision
-	// between a json.Marshal result and a stableStringify result is
-	// possible in theory; the prefix makes the source observable).
-	if !strings.HasPrefix(first, "__non_json:") {
-		t.Fatalf("canonicalSchemaString fallback must start with the __non_json: prefix, got: %s", first)
-	}
-}

internal/agent/prune.go

@@ -121,67 +121,3 @@ func toolNameForResult(messages []zeroruntime.Message, index int) string {
 func isPrunedPlaceholder(content string) bool {
 	return strings.HasPrefix(strings.TrimSpace(content), "[pruned ")
 }
-
-var supersedableReadTools = map[string]bool{
-	"read_file":          true,
-	"read_minified_file": true,
-	"grep":               true,
-	"glob":               true,
-	"list_directory":     true,
-}
-
-// pruneSupersededReadResults removes only older byte-identical results of the
-// same read call. It runs when context pressure already requires a rewrite, so
-// it does not trade a warm provider-cache prefix for a small apparent saving.
-func pruneSupersededReadResults(messages []zeroruntime.Message) ([]zeroruntime.Message, int) {
-	type seenResultKey struct {
-		call    string
-		content string
-	}
-	seen := map[seenResultKey]struct{}{}
-	out := messages
-	reclaimed := 0
-	copied := false
-
-	for index := len(messages) - 1; index >= 0; index-- {
-		message := messages[index]
-		if message.Role != zeroruntime.MessageRoleTool || isPrunedPlaceholder(message.Content) {
-			continue
-		}
-		call, ok := toolCallForResult(messages, index)
-		if !ok || !supersedableReadTools[call.Name] {
-			continue
-		}
-		key := seenResultKey{
-			call:    call.Name + "\x00" + strings.TrimSpace(call.Arguments),
-			content: message.Content,
-		}
-		if _, exists := seen[key]; exists {
-			placeholder := fmt.Sprintf("[superseded identical %s result — the newer result remains in context]", call.Name)
-			if !copied {
-				out = append([]zeroruntime.Message(nil), messages...)
-				copied = true
-			}
-			out[index].Content = placeholder
-			reclaimed += max(0, ApproxTextTokens(message.Content)-ApproxTextTokens(placeholder))
-			continue
-		}
-		seen[key] = struct{}{}
-	}
-	return out, reclaimed
-}
-
-func toolCallForResult(messages []zeroruntime.Message, resultIndex int) (zeroruntime.ToolCall, bool) {
-	id := messages[resultIndex].ToolCallID
-	if id == "" {
-		return zeroruntime.ToolCall{}, false
-	}
-	for index := resultIndex - 1; index >= 0; index-- {
-		for _, call := range messages[index].ToolCalls {
-			if call.ID == id {
-				return call, true
-			}
-		}
-	}
-	return zeroruntime.ToolCall{}, false
-}

internal/agent/prune_test.go

@@ -25,13 +25,6 @@ func toolCallMsg(id, name string) zeroruntime.Message {
 	}
 }
 
-func toolCallWithArgs(id, name, args string) zeroruntime.Message {
-	return zeroruntime.Message{
-		Role:      zeroruntime.MessageRoleAssistant,
-		ToolCalls: []zeroruntime.ToolCall{{ID: id, Name: name, Arguments: args}},
-	}
-}
-
 func TestPruneSkipsSmallSessions(t *testing.T) {
 	// Total reclaimable is under the gate → no pruning.
 	msgs := []zeroruntime.Message{
@@ -127,69 +120,3 @@ func TestPruneDoesNotMutateInput(t *testing.T) {
 		t.Fatal("pruneStaleToolOutput must not mutate the caller's slice")
 	}
 }
-
-func TestPruneSupersededReadResultsOnlyDropsIdenticalOlderReads(t *testing.T) {
-	body := strings.Repeat("same source line\n", 1000)
-	messages := []zeroruntime.Message{
-		toolCallWithArgs("old", "read_file", `{"path":"a.go"}`),
-		{Role: zeroruntime.MessageRoleTool, ToolCallID: "old", Content: body},
-		toolCallWithArgs("different", "read_file", `{"path":"b.go"}`),
-		{Role: zeroruntime.MessageRoleTool, ToolCallID: "different", Content: body},
-		toolCallWithArgs("new", "read_file", `{"path":"a.go"}`),
-		{Role: zeroruntime.MessageRoleTool, ToolCallID: "new", Content: body},
-	}
-	out, reclaimed := pruneSupersededReadResults(messages)
-	if reclaimed <= 0 || !strings.Contains(out[1].Content, "superseded identical") {
-		t.Fatalf("older duplicate was not pruned: reclaimed=%d output=%q", reclaimed, out[1].Content)
-	}
-	if out[3].Content != body || out[5].Content != body {
-		t.Fatal("different call and newest result must remain exact")
-	}
-	if messages[1].Content != body {
-		t.Fatal("input slice was mutated")
-	}
-}
-
-func TestPruneSupersededReadResultsKeepsChangedAndMutatingResults(t *testing.T) {
-	messages := []zeroruntime.Message{
-		toolCallWithArgs("r1", "read_file", `{"path":"a.go"}`),
-		{Role: zeroruntime.MessageRoleTool, ToolCallID: "r1", Content: "version one"},
-		toolCallWithArgs("r2", "read_file", `{"path":"a.go"}`),
-		{Role: zeroruntime.MessageRoleTool, ToolCallID: "r2", Content: "version two"},
-		toolCallWithArgs("e1", "edit_file", `{"path":"a.go"}`),
-		{Role: zeroruntime.MessageRoleTool, ToolCallID: "e1", Content: "edited"},
-		toolCallWithArgs("e2", "edit_file", `{"path":"a.go"}`),
-		{Role: zeroruntime.MessageRoleTool, ToolCallID: "e2", Content: "edited"},
-	}
-	out, reclaimed := pruneSupersededReadResults(messages)
-	if reclaimed != 0 {
-		t.Fatalf("unsafe results were pruned: %d", reclaimed)
-	}
-	for index := range messages {
-		if out[index].Content != messages[index].Content {
-			t.Fatalf("message %d changed unexpectedly", index)
-		}
-	}
-}
-
-func TestPruneSupersededReadResultsFindsMatchingBodyAcrossChangedRead(t *testing.T) {
-	bodyA := strings.Repeat("version a\n", 100)
-	bodyB := strings.Repeat("version b\n", 100)
-	messages := []zeroruntime.Message{
-		toolCallWithArgs("old-a", "read_file", `{"path":"a.go"}`),
-		{Role: zeroruntime.MessageRoleTool, ToolCallID: "old-a", Content: bodyA},
-		toolCallWithArgs("middle-b", "read_file", `{"path":"a.go"}`),
-		{Role: zeroruntime.MessageRoleTool, ToolCallID: "middle-b", Content: bodyB},
-		toolCallWithArgs("new-a", "read_file", `{"path":"a.go"}`),
-		{Role: zeroruntime.MessageRoleTool, ToolCallID: "new-a", Content: bodyA},
-	}
-
-	out, reclaimed := pruneSupersededReadResults(messages)
-
-	if reclaimed <= 0 || !strings.Contains(out[1].Content, "superseded identical") {
-		t.Fatalf("older matching body was not pruned across a changed result: reclaimed=%d output=%q", reclaimed, out[1].Content)
-	}
-	if out[3].Content != bodyB || out[5].Content != bodyA {
-		t.Fatal("changed middle result and newest matching result must remain exact")
-	}
-}

internal/agent/reconnect.go

@@ -8,7 +8,6 @@ import (
 	"time"
 
 	"github.com/Gitlawb/zero/internal/errhint"
-	"github.com/Gitlawb/zero/internal/trace"
 	"github.com/Gitlawb/zero/internal/zeroruntime"
 )
 
@@ -75,21 +74,14 @@ func stallRetryNoticeFor(options Options) reconnectNotifier {
 // already past its deadline is returned immediately (no retry) — those have
 // their own handling (compaction for context-limit, image-rejection, etc.).
 func streamWithReconnect(ctx context.Context, provider Provider, request zeroruntime.CompletionRequest, notify reconnectNotifier) (<-chan zeroruntime.StreamEvent, error) {
-	recorder := trace.FromContext(ctx)
 	stream, err := provider.StreamCompletion(ctx, request)
 	if err == nil {
-		if recorder != nil {
-			recorder.Counter(trace.CounterModelRequests, 1)
-		}
 		return stream, nil
 	}
 	for attempt := 1; attempt <= maxStreamReconnects; attempt++ {
 		if !shouldReconnect(ctx, err) {
 			return nil, err
 		}
-		if recorder != nil {
-			recorder.Counter(trace.CounterReconnectCount, 1)
-		}
 		if notify != nil {
 			notify(attempt, maxStreamReconnects)
 		}
@@ -98,9 +90,6 @@ func streamWithReconnect(ctx context.Context, provider Provider, request zerorun
 		}
 		stream, err = provider.StreamCompletion(ctx, request)
 		if err == nil {
-			if recorder != nil {
-				recorder.Counter(trace.CounterModelRequests, 1)
-			}
 			return stream, nil
 		}
 	}

internal/agent/request_permissions_test.go

@@ -180,7 +180,7 @@ func TestInlineAdditionalPermissionsRequiresPromptUnderAutoAllow(t *testing.T) {
 		},
 	}
 	decision := &sandbox.Decision{Action: sandbox.ActionAllow, AutoAllowed: true}
-	if !shouldRequestPermission(tools.NewScopedBashTool(t.TempDir(), nil), args, false, decision) {
+	if !shouldRequestPermission(tools.NewBashTool(t.TempDir()), args, false, decision) {
 		t.Fatal("inline additional permissions must request approval even when a sandboxed shell would otherwise auto-allow")
 	}
 }

internal/agent/selfcorrect.go

@@ -9,7 +9,6 @@ import (
 
 	"github.com/Gitlawb/zero/internal/lsp"
 	"github.com/Gitlawb/zero/internal/redaction"
-	"github.com/Gitlawb/zero/internal/trace"
 	"github.com/Gitlawb/zero/internal/verify"
 )
 
@@ -134,9 +133,7 @@ func (sc *SelfCorrector) inspect(ctx context.Context, changedFiles []string) Cor
 	}
 
 	if sc.cfg.IncludeTests && sc.verifier != nil {
-		verifySpan := trace.FromContext(ctx).Span(trace.SpanVerification)
 		vr, err := sc.verifier.Verify(ctx)
-		verifySpan.End()
 		if err != nil {
 			// "No plan / no tests" is not an error (DetectPlan returns an empty
 			// plan), so a non-nil error means verification could not run at all —

internal/agent/specialist_delegation_test.go

@@ -26,16 +26,6 @@ func TestSpecialistDelegationContext(t *testing.T) {
 	if strings.Contains(got, "nameless") {
 		t.Fatalf("nameless entry should be skipped: %q", got)
 	}
-	for _, want := range []string{"reduce total context", "Do not delegate small direct work", "bounded assignment"} {
-		if !strings.Contains(got, want) {
-			t.Fatalf("missing evidence-driven delegation guidance %q: %q", want, got)
-		}
-	}
-	for _, unwanted := range []string{"delegate to it proactively", "launch several specialists in parallel"} {
-		if strings.Contains(got, unwanted) {
-			t.Fatalf("retained broad delegation guidance %q: %q", unwanted, got)
-		}
-	}
 }
 
 func TestSystemPromptIncludesDelegationOnlyWithSpecialists(t *testing.T) {

internal/agent/system_prompt.go

@@ -11,7 +11,6 @@ import (
 
 	"github.com/Gitlawb/zero/internal/config"
 	"github.com/Gitlawb/zero/internal/repomap"
-	"github.com/Gitlawb/zero/internal/tools"
 	"github.com/Gitlawb/zero/internal/workspaceseed"
 )
 
@@ -71,22 +70,6 @@ const (
 // guidelines), and the safety confirmation policy. It is built once per run so
 // every turn shares one (cacheable) system turn.
 func buildSystemPrompt(options Options) string {
-	return buildSystemPromptParts(options).prompt
-}
-
-// systemPromptParts retains both the exact joined prompt and the diagnostic
-// sections used by prompt-prefix tracing. Building these together prevents the
-// tracing path from re-reading workspace state and guarantees that the hash is
-// computed from the same prompt bytes placed in the request.
-type systemPromptParts struct {
-	prompt             string
-	baseInstructions   string
-	confirmationPolicy string
-	projectContext     string
-	skills             string
-}
-
-func buildSystemPromptParts(options Options) systemPromptParts {
 	core := strings.TrimSpace(options.SystemPrompt)
 	if core == "" {
 		core = strings.TrimSpace(coreSystemPrompt)
@@ -95,9 +78,6 @@ func buildSystemPromptParts(options Options) systemPromptParts {
 		core = fallbackSystemPrompt
 	}
 	sections := []string{core}
-	if transient := strings.TrimSpace(options.TransientSystemPrompt); transient != "" {
-		sections = append(sections, transient)
-	}
 	if addendum := modelPromptAddendum(options.Model); addendum != "" {
 		sections = append(sections, addendum)
 	}
@@ -117,34 +97,22 @@ func buildSystemPromptParts(options Options) systemPromptParts {
 	if user := userGuidelines(); user != "" {
 		sections = append(sections, user)
 	}
-	project := workspaceContext(options.Cwd)
-	if project != "" {
-		sections = append(sections, project)
-	}
-	if modeCtx := permissionModeContext(options); modeCtx != "" {
-		sections = append(sections, modeCtx)
+	if ws := workspaceContext(options.Cwd); ws != "" {
+		sections = append(sections, ws)
 	}
 	if delegation := specialistDelegationContext(options); delegation != "" {
 		sections = append(sections, delegation)
 	}
-	skillsBlock := skillsContext(options)
-	if skillsBlock != "" {
+	if skillsBlock := skillsContext(options); skillsBlock != "" {
 		sections = append(sections, skillsBlock)
 	}
 	if style := responseStyleContext(options); style != "" {
 		sections = append(sections, style)
 	}
-	policy := strings.TrimSpace(confirmationPolicy)
-	if policy != "" {
+	if policy := strings.TrimSpace(confirmationPolicy); policy != "" {
 		sections = append(sections, policy)
 	}
-	return systemPromptParts{
-		prompt:             strings.Join(sections, "\n\n"),
-		baseInstructions:   core,
-		confirmationPolicy: policy,
-		projectContext:     project,
-		skills:             skillsBlock,
-	}
+	return strings.Join(sections, "\n\n")
 }
 
 // responseStyleContext renders the operator-selected reply style (TUI /style) as
@@ -185,19 +153,21 @@ func approvedCommandPrefixContext(options Options) string {
 	return "## Approved Command Prefixes\n\nThe following command prefixes have already been approved and do not need another permission prompt:\n" + strings.Join(lines, "\n")
 }
 
-// specialistDelegationContext describes when a specialist earns its extra
-// context and coordination cost. It renders only when specialists are known
-// (which is only where the Task tool is actually registered), so a run with no
-// delegatable specialists produces the previous prompt unchanged.
+// specialistDelegationContext nudges the orchestrator to offload read-heavy or
+// parallelizable work to a specialist sub-agent via the Task tool, keeping large
+// tool outputs out of the main context. It renders only when specialists are
+// known (which is only where the Task tool is actually registered), so a run with
+// no delegatable specialists produces the previous prompt unchanged.
 func specialistDelegationContext(options Options) string {
 	if len(options.Specialists) == 0 {
 		return ""
 	}
 	var b strings.Builder
 	b.WriteString("<specialists>\n")
-	b.WriteString("Use the Task tool when a specialist's focused context, expertise, or an independent parallel subtask is likely to improve quality or reduce total context. ")
-	b.WriteString("Do not delegate small direct work, duplicate exploration, or split sequential steps merely because a specialist is available. ")
-	b.WriteString("When delegating, give a bounded assignment and consume its concise evidence instead of importing a long transcript.\n")
+	b.WriteString("Delegate focused or read-heavy work to a specialist sub-agent with the Task tool instead of doing it inline. ")
+	b.WriteString("When a request matches a specialist's purpose, delegate to it proactively — you do not need the user to ask first. ")
+	b.WriteString("This keeps large tool outputs — searches, file dumps, multi-step exploration — out of your own context, so you stay fast and token-efficient. ")
+	b.WriteString("Prefer delegating codebase search and exploration; for independent subtasks, launch several specialists in parallel. Handle small, direct edits yourself.\n")
 	b.WriteString("Available specialists (call Task with the matching name when the task fits its purpose):\n")
 	for _, info := range options.Specialists {
 		name := strings.TrimSpace(info.Name)
@@ -315,13 +285,17 @@ func workspaceContext(cwd string) string {
 	b.WriteString("<environment>\n")
 	b.WriteString("Working directory: " + cwd + "\n")
 	b.WriteString("Operating system: " + runtime.GOOS + "\n")
-	b.WriteString(tools.HostShellEnvironmentGuidance() + "\n")
+	if runtime.GOOS == "windows" {
+		b.WriteString("Shell syntax: Windows cmd.exe syntax for exec_command/bash tools; prefer the workdir/cwd argument instead of cd when changing directories. Do not pipe to or invoke POSIX coreutils from Git for Windows (usr\\bin head/grep/tail/cat/...): they are MSYS binaries and fail under the write-restricted sandbox; use native Zero tools (grep, read_file, list_directory, glob) or cmd.exe findstr/more instead, or sandbox_permissions require_escalated only when host-level execution is truly required.\n")
+	} else {
+		b.WriteString("Shell syntax: /bin/sh syntax for exec_command/bash tools; prefer the workdir/cwd argument instead of cd when changing directories.\n")
+	}
 	if branch := gitBranchForPrompt(cwd); branch != "" {
 		b.WriteString("Git branch: " + branch + "\n")
 	}
 	b.WriteString("</environment>")
 
-	b.WriteString(projectGuidelines(cwd, FindProjectGitRoot(cwd)))
+	b.WriteString(projectGuidelines(cwd, findProjectGitRoot(cwd)))
 	if repoMap := repoMapContext(cwd); repoMap != "" {
 		b.WriteString("\n\n## Repo map\n\n" + repoMap)
 	}
@@ -568,17 +542,17 @@ func resolveDirCaseInsensitive(target, anchor string) (string, bool) {
 	return cur, true
 }
 
-// FindProjectGitRoot returns the nearest ancestor of cwd that contains a
+// findProjectGitRoot returns the nearest ancestor of cwd that contains a
 // .git entry (file or directory). Returns "" when no git root is found, so
 // the caller can fall back to cwd-only lookup.
-func FindProjectGitRoot(cwd string) string {
+func findProjectGitRoot(cwd string) string {
 	cwd = strings.TrimSpace(cwd)
 	if cwd == "" {
 		return ""
 	}
 	cur := cwd
 	for {
-		if HasGitMetadata(cur) {
+		if hasGitMetadata(cur) {
 			return cur
 		}
 		parent := filepath.Dir(cur)
@@ -589,7 +563,7 @@ func FindProjectGitRoot(cwd string) string {
 	}
 }
 
-func HasGitMetadata(dir string) bool {
+func hasGitMetadata(dir string) bool {
 	gitPath := filepath.Join(dir, ".git")
 	info, err := os.Stat(gitPath)
 	if err != nil {
@@ -645,11 +619,7 @@ func repoMapContext(cwd string) string {
 // cwd, handling both a regular checkout (.git dir) and a worktree (.git file).
 // Returns "" on any problem — the prompt simply omits the branch segment.
 func gitBranchForPrompt(cwd string) string {
-	gitRoot := FindProjectGitRoot(cwd)
-	if gitRoot == "" {
-		return ""
-	}
-	gitPath := filepath.Join(gitRoot, ".git")
+	gitPath := filepath.Join(cwd, ".git")
 	info, err := os.Stat(gitPath)
 	if err != nil {
 		return ""
@@ -665,10 +635,10 @@ func gitBranchForPrompt(cwd string) string {
 			return ""
 		}
 		// In worktree mode the gitdir is often RELATIVE (e.g.
-		// "gitdir: ../.git/worktrees/<name>") — resolve it against the worktree root (gitRoot), not the
+		// "gitdir: ../.git/worktrees/<name>") — resolve it against cwd, not the
 		// process working directory, or HEAD lookup fails and we drop the branch.
 		if !filepath.IsAbs(dir) {
-			dir = filepath.Join(gitRoot, dir)
+			dir = filepath.Join(cwd, dir)
 		}
 		headPath = filepath.Join(dir, "HEAD")
 	}
@@ -685,12 +655,3 @@ func gitBranchForPrompt(cwd string) string {
 	}
 	return ref
 }
-
-func permissionModeContext(options Options) string {
-	switch options.PermissionMode {
-	case PermissionModePlan:
-		return "Plan mode is active on this session. Your role is read-only exploration and planning: inspect the workspace and shape the plan with update_plan, but do not make changes to files or execute commands."
-	default:
-		return ""
-	}
-}

internal/agent/system_prompt.md

@@ -30,13 +30,16 @@ work.
    read-before-edit discipline: inspect the target file and nearby callers,
    tests, or config before you modify behavior. Never edit a file you have not
    read.
-2. **Plan.** Use update_plan for implementation or investigation that has at
-   least three meaningful dependent steps, then keep it current at milestones.
-   Mark completed work and the next in-progress step together when practical;
-   do not spend separate turns updating it after every file or command. Keep at
-   most one item in_progress. Skip plans for simple lookups, explanations, code
-   navigation, and other short tasks. Never create a plan after the work is
-   already complete merely to describe what you did.
+2. **Plan.** For multi-step work, call update_plan with an ordered checklist and
+   keep it live. The plan bar is the user's progress signal — call update_plan
+   after EACH concrete unit of work (every file written, every command run), not
+   just at coarse milestones: mark the finished item completed and the next one
+   in_progress before you start it. A plan stuck at 0/N while files are landing is
+   a bug, not economy — these calls are cheap, expected, and the update_plan cards
+   are hidden from the transcript, so frequent updates cost the user nothing and
+   never clutter the conversation. Keep at most one item in_progress, and never
+   batch the updates to the end of the turn. Skip the plan for trivial one-step
+   tasks.
 3. **Implement.** Make focused changes that match the surrounding code's style,
    naming, and conventions. Prefer the smallest change that fully solves the
    problem. Avoid broad refactors, unrelated rewrites, dependency churn, and
@@ -51,16 +54,11 @@ work.
 ## Editing discipline
 
 - Choose the narrowest tool that safely accomplishes the step. Prefer native
-  file tools - read_file, read_minified_file, list_directory, glob, grep,
-  write_file, edit_file, apply_patch - over shelling out to
-  cat/sed/awk/python for file operations.
+  file tools - read_file, list_directory, glob, grep, write_file, edit_file,
+  apply_patch - over shelling out to cat/sed/awk/python for file operations.
   They are safer, reviewable, and produce clean diffs.
-- Prefer read_minified_file when initially exploring source code; it preserves
-  code while removing comments and redundant whitespace. Use read_file when
-  exact text, comments, or line numbers are needed.
-- Keep edits focused and reviewable. A single patch may update several related
-  files when they form one coherent change; do not hide unrelated edits in a
-  bulk shell or script rewrite.
+- Make one tool call per file. Do not batch multi-file writes into a single
+  shell or script invocation.
 - For edits to existing files, prefer edit_file or apply_patch with minimal,
   targeted diffs. Match the existing indentation, imports, and idioms. Match the
   file's comment density: do not add explanatory comments unless the user asks or
@@ -102,9 +100,6 @@ work.
 - Use exec_command with `tty: true` for interactive terminal-style commands that
   need stdin beyond Ctrl-C. `/ps` and `/stop` are user-facing TUI commands; when
   you need to clean up a running foreground command yourself, use write_stdin.
-- write_stdin's session_id is only ever an id returned by a still-running
-  exec_command; never guess or probe ids. If you have no such session, start one
-  with exec_command, or use write_file/edit_file/apply_patch for file changes.
 - write_stdin with empty input polls an existing exec_command session, and
   `\u0003` interrupts it. Sending other stdin bytes may require approval because
   it can drive the running process beyond the original command. Non-tty sessions

internal/agent/system_prompt_models.go

@@ -71,6 +71,5 @@ const geminiPromptAddendum = `<model_guidance>
   equivalent shell commands; they are safer and produce cleaner diffs.
 - Be concise and concrete. When you run a shell command with side effects, state
   in one short clause why it is needed.
-- Use update_plan only for work with at least three meaningful dependent steps;
-  skip it for simple lookups and never create one after the work is complete.
+- Use update_plan for any multi-step task and keep it current.
 </model_guidance>`

internal/agent/system_prompt_test.go

@@ -51,16 +51,6 @@ func TestCoreSystemPromptIncludesCodingQualityRules(t *testing.T) {
 	}
 }
 
-func TestTransientSystemPromptIsIncludedOnlyWhenProvided(t *testing.T) {
-	const transient = "Handle this peer request under the current session policy."
-	if prompt := buildSystemPrompt(Options{}); strings.Contains(prompt, transient) {
-		t.Fatal("ordinary system prompt unexpectedly contains transient instructions")
-	}
-	if prompt := buildSystemPrompt(Options{TransientSystemPrompt: transient}); !strings.Contains(prompt, transient) {
-		t.Fatalf("transient instructions missing from system prompt:\n%s", prompt)
-	}
-}
-
 func TestSystemPromptExplainsSandboxEscalationForHiddenHostState(t *testing.T) {
 	prompt := strings.ToLower(buildSystemPrompt(Options{}))
 	for _, want := range []string{
@@ -434,7 +424,7 @@ func TestFindProjectGitRootIgnoresEmptyGitDirectory(t *testing.T) {
 	if err := os.MkdirAll(leaf, 0o755); err != nil {
 		t.Fatal(err)
 	}
-	if got := FindProjectGitRoot(leaf); got != "" {
-		t.Fatalf("FindProjectGitRoot = %q, want empty for invalid .git directory", got)
+	if got := findProjectGitRoot(leaf); got != "" {
+		t.Fatalf("findProjectGitRoot = %q, want empty for invalid .git directory", got)
 	}
 }

internal/agent/task_state.go

@@ -1,566 +0,0 @@
-package agent
-
-import (
-	"crypto/sha256"
-	"encoding/json"
-	"fmt"
-	"reflect"
-	"regexp"
-	"sort"
-	"strings"
-	"unicode/utf8"
-
-	"github.com/Gitlawb/zero/internal/tools"
-	"github.com/Gitlawb/zero/internal/trace"
-	"github.com/Gitlawb/zero/internal/zeroruntime"
-)
-
-type taskStatus string
-
-const (
-	taskStatusActive     taskStatus = "active"
-	taskStatusComplete   taskStatus = "complete"
-	taskStatusIncomplete taskStatus = "incomplete"
-)
-
-type taskPlanParity string
-
-const (
-	taskPlanParityUnknown  taskPlanParity = "unknown"
-	taskPlanParityMatch    taskPlanParity = "match"
-	taskPlanParityMismatch taskPlanParity = "mismatch"
-)
-
-type taskPlanItem struct {
-	Content string `json:"content"`
-	Status  string `json:"status"`
-	Notes   string `json:"notes,omitempty"`
-}
-
-type taskPlanState struct {
-	Items      []taskPlanItem `json:"items,omitempty"`
-	Pending    int            `json:"pending"`
-	InProgress int            `json:"in_progress"`
-	Completed  int            `json:"completed"`
-	Failed     int            `json:"failed"`
-}
-
-type taskToolState struct {
-	Succeeded int `json:"succeeded"`
-	Failed    int `json:"failed"`
-}
-
-type taskVerificationState struct {
-	Passed      int     `json:"passed"`
-	Failed      int     `json:"failed"`
-	LastOutcome Outcome `json:"last_outcome,omitempty"`
-}
-
-type taskFailureState struct {
-	Key     string `json:"key,omitempty"`
-	Tool    string `json:"tool"`
-	Command string `json:"command,omitempty"`
-	Summary string `json:"summary"`
-}
-
-type taskApprovalState struct {
-	Tool     string                   `json:"tool"`
-	Decision PermissionDecisionAction `json:"decision,omitempty"`
-	Scope    string                   `json:"scope,omitempty"`
-	Reason   string                   `json:"reason,omitempty"`
-}
-
-type taskArtifactState struct {
-	Tool string `json:"tool"`
-	Path string `json:"path"`
-}
-
-type taskStateSnapshot struct {
-	Revision            int                   `json:"revision"`
-	Objective           string                `json:"objective"`
-	Status              taskStatus            `json:"status"`
-	Plan                taskPlanState         `json:"plan"`
-	Tools               taskToolState         `json:"tools"`
-	Verification        taskVerificationState `json:"verification"`
-	ChangedFiles        []string              `json:"changed_files,omitempty"`
-	Constraints         []string              `json:"constraints,omitempty"`
-	UnresolvedFailures  []taskFailureState    `json:"unresolved_failures,omitempty"`
-	Approvals           []taskApprovalState   `json:"approvals,omitempty"`
-	Artifacts           []taskArtifactState   `json:"artifacts,omitempty"`
-	ResolvedFailureKeys []string              `json:"-"`
-	CompletionDecision  CompletionDecision    `json:"completion_decision,omitempty"`
-	CompletionReason    string                `json:"completion_reason,omitempty"`
-	PlanParity          taskPlanParity        `json:"plan_parity"`
-}
-
-type taskStateEventKind int
-
-const (
-	taskStateEventPlan taskStateEventKind = iota + 1
-	taskStateEventToolResult
-	taskStateEventVerification
-	taskStateEventCompletion
-	taskStateEventPermission
-)
-
-type taskStateEvent struct {
-	kind         taskStateEventKind
-	arguments    string
-	toolResult   ToolResult
-	verification Outcome
-	completion   completionEvaluation
-	permission   PermissionEvent
-}
-
-// taskState is a deterministic projection of facts the loop already observes.
-// It does not initiate work or replace the transcript; consumers must check
-// transcript parity before using its compact representation.
-type taskState struct {
-	snapshotValue taskStateSnapshot
-	changedFiles  map[string]struct{}
-	recorder      *trace.Recorder
-	planObserved  bool
-}
-
-func newTaskState(objective string, recorder *trace.Recorder) *taskState {
-	state := &taskState{
-		snapshotValue: taskStateSnapshot{
-			Objective:  strings.TrimSpace(objective),
-			Status:     taskStatusActive,
-			PlanParity: taskPlanParityUnknown,
-		},
-		changedFiles: map[string]struct{}{},
-	}
-	state.snapshotValue.Constraints = extractExplicitConstraints(objective)
-	state.recorder = recorder
-	return state
-}
-
-func (state *taskState) observe(event taskStateEvent) bool {
-	if state == nil {
-		return false
-	}
-	changed := false
-	switch event.kind {
-	case taskStateEventPlan:
-		plan, ok := parseTaskPlan(event.arguments)
-		if !ok {
-			return false
-		}
-		state.snapshotValue.Plan = summarizeTaskPlan(plan)
-		state.planObserved = true
-		state.markActive()
-		changed = true
-	case taskStateEventToolResult:
-		state.markActive()
-		if event.toolResult.Status == tools.StatusOK {
-			state.snapshotValue.Tools.Succeeded++
-		} else {
-			state.snapshotValue.Tools.Failed++
-		}
-		for _, path := range event.toolResult.ChangedFiles {
-			path = strings.TrimSpace(path)
-			if path != "" {
-				state.changedFiles[path] = struct{}{}
-			}
-		}
-		state.snapshotValue.ChangedFiles = sortedKeys(state.changedFiles)
-		state.observeToolEvidence(event.arguments, event.toolResult)
-		changed = true
-	case taskStateEventVerification:
-		if event.verification == OutcomeDisabled {
-			return false
-		}
-		state.markActive()
-		state.snapshotValue.Verification.LastOutcome = event.verification
-		switch event.verification {
-		case OutcomePassed:
-			state.snapshotValue.Verification.Passed++
-		case OutcomeCorrecting, OutcomeReported, OutcomeAborted:
-			state.snapshotValue.Verification.Failed++
-		}
-		changed = true
-	case taskStateEventCompletion:
-		state.snapshotValue.CompletionDecision = event.completion.Decision
-		state.snapshotValue.CompletionReason = event.completion.Reason
-		switch event.completion.Decision {
-		case CompletionComplete:
-			state.snapshotValue.Status = taskStatusComplete
-		case CompletionIncomplete:
-			state.snapshotValue.Status = taskStatusIncomplete
-		default:
-			state.snapshotValue.Status = taskStatusActive
-		}
-		changed = true
-	case taskStateEventPermission:
-		state.markActive()
-		state.observePermission(event.permission)
-		changed = true
-	}
-	if changed {
-		state.snapshotValue.Revision++
-		if event.kind != taskStateEventToolResult {
-			state.emit()
-		}
-	}
-	return changed
-}
-
-func (state *taskState) markActive() {
-	state.snapshotValue.Status = taskStatusActive
-	state.snapshotValue.CompletionDecision = ""
-	state.snapshotValue.CompletionReason = ""
-}
-
-func (state *taskState) snapshot() taskStateSnapshot {
-	if state == nil {
-		return taskStateSnapshot{}
-	}
-	snapshot := state.snapshotValue
-	snapshot.Plan.Items = append([]taskPlanItem(nil), state.snapshotValue.Plan.Items...)
-	snapshot.ChangedFiles = append([]string(nil), state.snapshotValue.ChangedFiles...)
-	snapshot.Constraints = append([]string(nil), state.snapshotValue.Constraints...)
-	snapshot.UnresolvedFailures = append([]taskFailureState(nil), state.snapshotValue.UnresolvedFailures...)
-	snapshot.Approvals = append([]taskApprovalState(nil), state.snapshotValue.Approvals...)
-	snapshot.Artifacts = append([]taskArtifactState(nil), state.snapshotValue.Artifacts...)
-	snapshot.ResolvedFailureKeys = append([]string(nil), state.snapshotValue.ResolvedFailureKeys...)
-	return snapshot
-}
-
-const (
-	maxTaskEvidenceEntries = 8
-	maxTaskEvidenceBytes   = 320
-	maxTaskConstraints     = 10
-)
-
-func (state *taskState) observeToolEvidence(arguments string, result ToolResult) {
-	command := capTaskEvidence(commandFromArguments(arguments))
-	key := taskFailureKey(result.Name, command, arguments)
-	if result.Status == tools.StatusOK {
-		if hasTaskFailure(state.snapshotValue.UnresolvedFailures, key) {
-			state.snapshotValue.UnresolvedFailures = removeTaskFailure(state.snapshotValue.UnresolvedFailures, key)
-			state.snapshotValue.ResolvedFailureKeys = appendBoundedUnique(state.snapshotValue.ResolvedFailureKeys, key, maxTaskEvidenceEntries)
-		}
-	} else {
-		state.snapshotValue.ResolvedFailureKeys = removeString(state.snapshotValue.ResolvedFailureKeys, key)
-		state.snapshotValue.UnresolvedFailures = removeTaskFailure(state.snapshotValue.UnresolvedFailures, key)
-		state.snapshotValue.UnresolvedFailures = appendBounded(state.snapshotValue.UnresolvedFailures, taskFailureState{
-			Key: key, Tool: result.Name, Command: command, Summary: capTaskEvidence(firstLine(result.Output)),
-		}, maxTaskEvidenceEntries)
-	}
-	for _, metaKey := range []string{"spill_path", "artifact_path"} {
-		if path := strings.TrimSpace(result.Meta[metaKey]); path != "" {
-			state.snapshotValue.Artifacts = appendBoundedUnique(state.snapshotValue.Artifacts,
-				taskArtifactState{Tool: result.Name, Path: capTaskEvidence(path)}, maxTaskEvidenceEntries)
-		}
-	}
-}
-
-func taskFailureKey(toolName, command, arguments string) string {
-	identity := command
-	if identity == "" {
-		canonical := []byte(strings.TrimSpace(arguments))
-		var value any
-		if json.Unmarshal(canonical, &value) == nil {
-			if encoded, err := json.Marshal(value); err == nil {
-				canonical = encoded
-			}
-		}
-		digest := sha256.Sum256(canonical)
-		identity = fmt.Sprintf("args:%x", digest[:8])
-	}
-	return strings.TrimSpace(toolName) + "\x00" + identity
-}
-
-func (state *taskState) observePermission(event PermissionEvent) {
-	decision := event.DecisionAction
-	if decision == "" {
-		decision = PermissionDecisionAction(event.Action)
-	}
-	state.snapshotValue.Approvals = appendBoundedUnique(state.snapshotValue.Approvals, taskApprovalState{
-		Tool: event.ToolName, Decision: decision, Scope: capTaskEvidence(event.Scope),
-		Reason: capTaskEvidence(firstNonEmptyString(event.DecisionReason, event.Reason)),
-	}, maxTaskEvidenceEntries)
-}
-
-func commandFromArguments(arguments string) string {
-	var object map[string]json.RawMessage
-	if json.Unmarshal([]byte(strings.TrimSpace(arguments)), &object) != nil {
-		return ""
-	}
-	for _, key := range []string{"cmd", "command", "script", "shell"} {
-		var command string
-		if raw, ok := object[key]; ok && json.Unmarshal(raw, &command) == nil {
-			return strings.TrimSpace(command)
-		}
-	}
-	return ""
-}
-
-func firstLine(value string) string {
-	value = strings.TrimSpace(value)
-	if index := strings.IndexByte(value, '\n'); index >= 0 {
-		value = value[:index]
-	}
-	return value
-}
-
-func capTaskEvidence(value string) string {
-	value = strings.TrimSpace(value)
-	if len(value) <= maxTaskEvidenceBytes {
-		return value
-	}
-	limit := maxTaskEvidenceBytes - 3
-	for limit > 0 && !utf8.RuneStart(value[limit]) {
-		limit--
-	}
-	return strings.TrimSpace(value[:limit]) + "..."
-}
-
-var explicitConstraintPatterns = []*regexp.Regexp{
-	regexp.MustCompile(`(?i)\bprefer(?:s|red|ring)?\s+\w`),
-	regexp.MustCompile(`(?i)\bdon'?t want\b`),
-	regexp.MustCompile(`(?i)\balways (?:use|do|run|prefer|keep|make|format|write|add|set|put|prefix|start|include|append)\b`),
-	regexp.MustCompile(`(?i)\bnever (?:use|do|run|push|commit|write|ignore|add|set|put|remove|delete|include|deploy)\b`),
-	regexp.MustCompile(`(?i)\bplease (?:use|avoid|keep|make|don'?t|do not|format|write)\b`),
-	regexp.MustCompile(`(?i)\b(?:style|format|language|naming)\s*[:=]\s*\S`),
-}
-
-func extractExplicitConstraints(text string) []string {
-	var constraints []string
-	for _, line := range strings.Split(text, "\n") {
-		line = strings.TrimSpace(line)
-		if len(line) < 5 || len(line) > 200 || strings.HasSuffix(line, "?") || strings.Contains(line, "?...") {
-			continue
-		}
-		for _, pattern := range explicitConstraintPatterns {
-			if pattern.MatchString(line) {
-				constraints = appendBoundedUnique(constraints, line, maxTaskConstraints)
-				break
-			}
-		}
-		if len(constraints) == maxTaskConstraints {
-			break
-		}
-	}
-	return constraints
-}
-
-func appendBounded[T any](values []T, value T, limit int) []T {
-	values = append(values, value)
-	if len(values) > limit {
-		values = values[len(values)-limit:]
-	}
-	return values
-}
-
-func appendBoundedUnique[T comparable](values []T, value T, limit int) []T {
-	filtered := values[:0]
-	for _, existing := range values {
-		if existing != value {
-			filtered = append(filtered, existing)
-		}
-	}
-	return appendBounded(filtered, value, limit)
-}
-
-func hasTaskFailure(values []taskFailureState, key string) bool {
-	for _, value := range values {
-		if value.Key == key {
-			return true
-		}
-	}
-	return false
-}
-
-func removeTaskFailure(values []taskFailureState, key string) []taskFailureState {
-	out := values[:0]
-	for _, value := range values {
-		if value.Key != key {
-			out = append(out, value)
-		}
-	}
-	return out
-}
-
-func removeString(values []string, target string) []string {
-	out := make([]string, 0, len(values))
-	for _, value := range values {
-		if value != target {
-			out = append(out, value)
-		}
-	}
-	return out
-}
-
-// observePlanParity compares only the latest plan projection with the plan tool
-// calls still present in messages. It mutates the snapshot and emits when the
-// parity value changes; objective, tool, and verification fields are not part of
-// this comparison.
-func (state *taskState) observePlanParity(messages []zeroruntime.Message) taskPlanParity {
-	if state == nil {
-		return taskPlanParityUnknown
-	}
-	transcriptPlan, found, valid := latestTaskPlan(messages)
-	tracked := state.snapshotValue.Plan.Items
-	parity := taskPlanParityMatch
-	if !valid || found != state.planObserved || (found && !reflect.DeepEqual(transcriptPlan, tracked)) {
-		parity = taskPlanParityMismatch
-	}
-	if state.snapshotValue.PlanParity != parity {
-		state.snapshotValue.PlanParity = parity
-		state.snapshotValue.Revision++
-		state.emit()
-	}
-	return parity
-}
-
-type completionContext struct {
-	Objective             string
-	PlanPending           bool
-	PlanMatchesTranscript bool
-}
-
-func (state *taskState) completionContext(messages []zeroruntime.Message, transcriptPlanPending bool) completionContext {
-	context := completionContext{PlanPending: transcriptPlanPending}
-	if state == nil {
-		return context
-	}
-	context.Objective = state.snapshotValue.Objective
-	context.PlanMatchesTranscript = state.observePlanParity(messages) == taskPlanParityMatch
-	if context.PlanMatchesTranscript {
-		context.PlanPending = state.snapshotValue.Plan.Pending+state.snapshotValue.Plan.InProgress > 0
-	}
-	return context
-}
-
-func (state *taskState) snapshotForCompaction(messages []zeroruntime.Message) *taskStateSnapshot {
-	if state == nil {
-		return nil
-	}
-	state.observePlanParity(messages)
-	snapshot := state.snapshot()
-	return &snapshot
-}
-
-func (state *taskState) emit() {
-	if state == nil || state.recorder == nil {
-		return
-	}
-	snapshot := state.snapshotValue
-	state.recorder.EmitTaskState(trace.TaskStateEvent{
-		Revision:            snapshot.Revision,
-		Status:              string(snapshot.Status),
-		PlanPending:         snapshot.Plan.Pending,
-		PlanInProgress:      snapshot.Plan.InProgress,
-		PlanCompleted:       snapshot.Plan.Completed,
-		PlanFailed:          snapshot.Plan.Failed,
-		ToolsSucceeded:      snapshot.Tools.Succeeded,
-		ToolsFailed:         snapshot.Tools.Failed,
-		VerificationPassed:  snapshot.Verification.Passed,
-		VerificationFailed:  snapshot.Verification.Failed,
-		VerificationOutcome: string(snapshot.Verification.LastOutcome),
-		ChangedFileCount:    len(snapshot.ChangedFiles),
-		CompletionDecision:  string(snapshot.CompletionDecision),
-		PlanParity:          string(snapshot.PlanParity),
-	})
-}
-
-func parseTaskPlan(arguments string) ([]taskPlanItem, bool) {
-	var object map[string]json.RawMessage
-	if json.Unmarshal([]byte(strings.TrimSpace(arguments)), &object) != nil {
-		return nil, false
-	}
-	rawPlan, ok := object["plan"]
-	if !ok || string(rawPlan) == "null" {
-		return nil, false
-	}
-	var parsed []struct {
-		ID      *string `json:"id"`
-		Content string  `json:"content"`
-		Status  string  `json:"status"`
-		Notes   string  `json:"notes"`
-	}
-	if json.Unmarshal(rawPlan, &parsed) != nil {
-		return nil, false
-	}
-	plan := make([]taskPlanItem, 0, len(parsed))
-	for _, raw := range parsed {
-		if raw.Content == "" {
-			return nil, false
-		}
-		plan = append(plan, taskPlanItem{
-			Content: raw.Content,
-			Status:  normalizeTaskPlanStatus(raw.Status),
-			Notes:   raw.Notes,
-		})
-	}
-	lastInProgress := -1
-	for index := range plan {
-		if plan[index].Status == "in_progress" {
-			if lastInProgress >= 0 {
-				plan[lastInProgress].Status = "completed"
-			}
-			lastInProgress = index
-		}
-	}
-	if len(plan) == 0 {
-		return nil, true
-	}
-	return plan, true
-}
-
-func normalizeTaskPlanStatus(status string) string {
-	switch strings.ToLower(strings.TrimSpace(status)) {
-	case "completed", "complete", "done", "finished", "resolved", "✓", "x", "[x]":
-		return "completed"
-	case "in_progress", "in-progress", "inprogress", "in progress", "active", "doing", "started", "current", "wip", "ongoing", "running":
-		return "in_progress"
-	case "failed", "fail", "error", "errored", "blocked", "cancelled", "canceled", "abandoned", "skipped":
-		return "failed"
-	default:
-		return "pending"
-	}
-}
-
-func summarizeTaskPlan(items []taskPlanItem) taskPlanState {
-	plan := taskPlanState{Items: append([]taskPlanItem(nil), items...)}
-	for _, item := range items {
-		switch item.Status {
-		case "completed":
-			plan.Completed++
-		case "in_progress":
-			plan.InProgress++
-		case "failed":
-			plan.Failed++
-		default:
-			plan.Pending++
-		}
-	}
-	return plan
-}
-
-func latestTaskPlan(messages []zeroruntime.Message) ([]taskPlanItem, bool, bool) {
-	for i := len(messages) - 1; i >= 0; i-- {
-		for j := len(messages[i].ToolCalls) - 1; j >= 0; j-- {
-			call := messages[i].ToolCalls[j]
-			if call.Name != planToolName {
-				continue
-			}
-			plan, ok := parseTaskPlan(call.Arguments)
-			return plan, true, ok
-		}
-	}
-	return nil, false, true
-}
-
-func sortedKeys(values map[string]struct{}) []string {
-	keys := make([]string, 0, len(values))
-	for value := range values {
-		keys = append(keys, value)
-	}
-	// Paths are evidence, not chronology. Stable ordering makes replay snapshots
-	// deterministic even if events originate from parallel read batches later.
-	sort.Strings(keys)
-	return keys
-}

internal/agent/task_state_test.go

@@ -1,266 +0,0 @@
-package agent
-
-import (
-	"context"
-	"reflect"
-	"testing"
-
-	"github.com/Gitlawb/zero/internal/tools"
-	"github.com/Gitlawb/zero/internal/trace"
-	"github.com/Gitlawb/zero/internal/zeroruntime"
-)
-
-func TestTaskStateReplayIsDeterministic(t *testing.T) {
-	events := []taskStateEvent{
-		{kind: taskStateEventPlan, arguments: `{"plan":[{"content":"inspect","status":"done"},{"content":"implement","status":"in progress"},{"content":"verify"}]}`},
-		{kind: taskStateEventToolResult, toolResult: ToolResult{Name: "apply_patch", Status: tools.StatusOK, ChangedFiles: []string{"a.go", "a.go", "b.go"}}},
-		{kind: taskStateEventToolResult, toolResult: ToolResult{Name: "go test", Status: tools.StatusError}},
-		{kind: taskStateEventVerification, verification: OutcomePassed},
-		{kind: taskStateEventCompletion, completion: completionEvaluation{Decision: CompletionUncertain, Reason: "pending work"}},
-	}
-
-	first := newTaskState("ship the change", nil)
-	second := newTaskState("ship the change", nil)
-	for _, event := range events {
-		first.observe(event)
-		second.observe(event)
-	}
-
-	if !reflect.DeepEqual(first.snapshot(), second.snapshot()) {
-		t.Fatalf("replaying the same events produced different snapshots:\nfirst:  %#v\nsecond: %#v", first.snapshot(), second.snapshot())
-	}
-	got := first.snapshot()
-	if got.Status != taskStatusActive || got.Plan.Pending != 1 || got.Plan.InProgress != 1 || got.Plan.Completed != 1 {
-		t.Fatalf("unexpected plan snapshot: %#v", got)
-	}
-	if got.Tools.Succeeded != 1 || got.Tools.Failed != 1 || len(got.ChangedFiles) != 2 {
-		t.Fatalf("unexpected evidence snapshot: %#v", got)
-	}
-	if got.Verification.Passed != 1 || got.Verification.Failed != 0 || got.Verification.LastOutcome != OutcomePassed {
-		t.Fatalf("unexpected verification snapshot: %#v", got.Verification)
-	}
-}
-
-func TestTaskStateCoalescesToolResultsIntoNextTraceEvent(t *testing.T) {
-	recorder := trace.NewRecorder("session", "run", "")
-	state := newTaskState("objective", recorder)
-	for range 20 {
-		state.observe(taskStateEvent{kind: taskStateEventToolResult, toolResult: ToolResult{Status: tools.StatusOK}})
-	}
-	if events := recorder.Finish().TaskStates; len(events) != 0 {
-		t.Fatalf("tool results should be coalesced, got %d trace events", len(events))
-	}
-
-	recorder = trace.NewRecorder("session", "run-2", "")
-	state = newTaskState("objective", recorder)
-	for range 20 {
-		state.observe(taskStateEvent{kind: taskStateEventToolResult, toolResult: ToolResult{Status: tools.StatusOK}})
-	}
-	state.observe(taskStateEvent{kind: taskStateEventCompletion, completion: completionEvaluation{Decision: CompletionComplete}})
-	events := recorder.Finish().TaskStates
-	if len(events) != 1 || events[0].ToolsSucceeded != 20 {
-		t.Fatalf("coalesced tool total missing from completion snapshot: %#v", events)
-	}
-}
-
-func TestRunEmitsTaskStateFromExistingLoopEvents(t *testing.T) {
-	registry := tools.NewRegistry()
-	registry.Register(tools.NewUpdatePlanTool())
-	provider := &mockProvider{turns: [][]zeroruntime.StreamEvent{
-		{
-			{Type: zeroruntime.StreamEventToolCallStart, ToolCallID: "plan-1", ToolName: planToolName},
-			{Type: zeroruntime.StreamEventToolCallDelta, ToolCallID: "plan-1", ArgumentsFragment: `{"plan":[{"content":"implement","status":"completed"}]}`},
-			{Type: zeroruntime.StreamEventToolCallEnd, ToolCallID: "plan-1"},
-			{Type: zeroruntime.StreamEventDone},
-		},
-		{
-			{Type: zeroruntime.StreamEventText, Content: "done"},
-			{Type: zeroruntime.StreamEventDone},
-		},
-	}}
-	recorder := trace.NewRecorder("session", "run", "")
-
-	result, err := Run(context.Background(), "implement the change", provider, Options{
-		Registry: registry,
-		Trace:    recorder,
-	})
-	if err != nil {
-		t.Fatalf("Run: %v", err)
-	}
-	if result.FinalAnswer != "done" {
-		t.Fatalf("final answer = %q, want done", result.FinalAnswer)
-	}
-	events := recorder.Finish().TaskStates
-	if len(events) < 3 {
-		t.Fatalf("expected plan, tool, completion, and parity snapshots, got %#v", events)
-	}
-	last := events[len(events)-1]
-	if last.Status != string(taskStatusComplete) || last.PlanCompleted != 1 || last.ToolsSucceeded != 1 || last.PlanParity != string(taskPlanParityMatch) {
-		t.Fatalf("unexpected final task snapshot: %#v", last)
-	}
-}
-
-func TestTaskStateSnapshotIsImmutable(t *testing.T) {
-	state := newTaskState("objective", nil)
-	state.observe(taskStateEvent{kind: taskStateEventPlan, arguments: `{"plan":[{"content":"one","status":"pending"}]}`})
-	state.observe(taskStateEvent{kind: taskStateEventToolResult, toolResult: ToolResult{Status: tools.StatusOK, ChangedFiles: []string{"one.go"}}})
-
-	snapshot := state.snapshot()
-	snapshot.Plan.Items[0].Content = "mutated"
-	snapshot.ChangedFiles[0] = "mutated.go"
-
-	fresh := state.snapshot()
-	if fresh.Plan.Items[0].Content != "one" || fresh.ChangedFiles[0] != "one.go" {
-		t.Fatalf("snapshot mutation leaked into task state: %#v", fresh)
-	}
-}
-
-func TestTaskStateRecordsDurableRuntimeEvidence(t *testing.T) {
-	state := newTaskState("Please keep the change focused.\nNever commit generated reports.", nil)
-	arguments := `{"cmd":"go test ./internal/agent"}`
-	state.observe(taskStateEvent{kind: taskStateEventToolResult, arguments: arguments, toolResult: ToolResult{
-		ToolCallID: "test-1", Name: "exec_command", Status: tools.StatusError,
-		Output: "Error: TestResumeRetainsState failed\nfull diagnostic",
-		Meta:   map[string]string{"spill_path": ".zero/artifacts/test-1.txt"},
-	}})
-	state.observe(taskStateEvent{kind: taskStateEventPermission, permission: PermissionEvent{
-		ToolName: "exec_command", DecisionAction: PermissionDecisionAllowForSession,
-		Scope: "/tmp", DecisionReason: "write a temporary fixture",
-	}})
-
-	snapshot := state.snapshot()
-	if !reflect.DeepEqual(snapshot.Constraints, []string{"Please keep the change focused.", "Never commit generated reports."}) {
-		t.Fatalf("unexpected explicit constraints: %#v", snapshot.Constraints)
-	}
-	if len(snapshot.UnresolvedFailures) != 1 || snapshot.UnresolvedFailures[0].Command != "go test ./internal/agent" || snapshot.UnresolvedFailures[0].Summary != "Error: TestResumeRetainsState failed" {
-		t.Fatalf("unexpected failure evidence: %#v", snapshot.UnresolvedFailures)
-	}
-	if len(snapshot.Artifacts) != 1 || snapshot.Artifacts[0].Path != ".zero/artifacts/test-1.txt" {
-		t.Fatalf("unexpected artifact evidence: %#v", snapshot.Artifacts)
-	}
-	if len(snapshot.Approvals) != 1 || snapshot.Approvals[0].Decision != PermissionDecisionAllowForSession || snapshot.Approvals[0].Scope != "/tmp" {
-		t.Fatalf("unexpected approval evidence: %#v", snapshot.Approvals)
-	}
-	state.observe(taskStateEvent{kind: taskStateEventToolResult, arguments: `{"path":"a.go"}`, toolResult: ToolResult{
-		ToolCallID: "read-a", Name: "read_file", Status: tools.StatusError, Output: "Error: a.go is unavailable",
-	}})
-	state.observe(taskStateEvent{kind: taskStateEventToolResult, arguments: `{"path":"b.go"}`, toolResult: ToolResult{
-		ToolCallID: "read-b", Name: "read_file", Status: tools.StatusOK, Output: "package b",
-	}})
-	if failures := state.snapshot().UnresolvedFailures; len(failures) != 2 || failures[1].Summary != "Error: a.go is unavailable" {
-		t.Fatalf("success for a different target resolved the wrong failure: %#v", failures)
-	}
-
-	state.observe(taskStateEvent{kind: taskStateEventToolResult, arguments: arguments, toolResult: ToolResult{
-		ToolCallID: "test-2", Name: "exec_command", Status: tools.StatusOK, Output: "ok",
-	}})
-	if failures := state.snapshot().UnresolvedFailures; len(failures) != 1 || failures[0].Summary != "Error: a.go is unavailable" {
-		t.Fatalf("successful retry did not resolve only its matching failure: %#v", failures)
-	}
-	state.observe(taskStateEvent{kind: taskStateEventToolResult, arguments: `{"path":"a.go"}`, toolResult: ToolResult{
-		ToolCallID: "read-a-2", Name: "read_file", Status: tools.StatusOK, Output: "package a",
-	}})
-	if failures := state.snapshot().UnresolvedFailures; len(failures) != 0 {
-		t.Fatalf("successful non-command retry did not resolve its matching failure: %#v", failures)
-	}
-	state.observe(taskStateEvent{kind: taskStateEventToolResult, arguments: `{"path":"a.go"}`, toolResult: ToolResult{
-		ToolCallID: "read-a-3", Name: "read_file", Status: tools.StatusError, Output: "Error: a.go failed again",
-	}})
-	snapshot = state.snapshot()
-	if len(snapshot.UnresolvedFailures) != 1 || snapshot.UnresolvedFailures[0].Summary != "Error: a.go failed again" {
-		t.Fatalf("a failure recurring after success was hidden: %#v", snapshot.UnresolvedFailures)
-	}
-	for _, key := range snapshot.ResolvedFailureKeys {
-		if key == snapshot.UnresolvedFailures[0].Key {
-			t.Fatalf("recurring failure retained a stale resolved key: %#v", snapshot.ResolvedFailureKeys)
-		}
-	}
-}
-
-func TestTaskStatePlanParityUsesLatestPlan(t *testing.T) {
-	state := newTaskState("objective", nil)
-	state.observe(taskStateEvent{kind: taskStateEventPlan, arguments: `{"plan":[{"content":"old","status":"completed"}]}`})
-	state.observe(taskStateEvent{kind: taskStateEventPlan, arguments: `{"plan":[{"content":"new","status":"in_progress"}]}`})
-
-	messages := []zeroruntime.Message{{Role: zeroruntime.MessageRoleAssistant, ToolCalls: []zeroruntime.ToolCall{
-		{Name: planToolName, Arguments: `{"plan":[{"content":"new","status":"in_progress"}]}`},
-	}}}
-	if parity := state.observePlanParity(messages); parity != taskPlanParityMatch {
-		t.Fatalf("parity = %q, want match", parity)
-	}
-
-	messages[0].ToolCalls[0].Arguments = `{"plan":[{"content":"different","status":"pending"}]}`
-	if parity := state.observePlanParity(messages); parity != taskPlanParityMismatch {
-		t.Fatalf("parity = %q, want mismatch", parity)
-	}
-}
-
-func TestTaskStateMatchesPlanToolNormalization(t *testing.T) {
-	state := newTaskState("objective", nil)
-	arguments := `{"plan":[{"content":"first","status":"in_progress"},{"content":"second","status":"in_progress"}]}`
-	state.observe(taskStateEvent{kind: taskStateEventPlan, arguments: arguments})
-
-	snapshot := state.snapshot()
-	if snapshot.Plan.Completed != 1 || snapshot.Plan.InProgress != 1 {
-		t.Fatalf("multiple active items were not normalized like the plan tool: %#v", snapshot.Plan)
-	}
-	messages := []zeroruntime.Message{{Role: zeroruntime.MessageRoleAssistant, ToolCalls: []zeroruntime.ToolCall{{Name: planToolName, Arguments: arguments}}}}
-	if parity := state.observePlanParity(messages); parity != taskPlanParityMatch {
-		t.Fatalf("normalized state should still match its transcript event, got %q", parity)
-	}
-
-	empty := newTaskState("objective", nil)
-	empty.observe(taskStateEvent{kind: taskStateEventPlan, arguments: `{"plan":[]}`})
-	emptyMessages := []zeroruntime.Message{{Role: zeroruntime.MessageRoleAssistant, ToolCalls: []zeroruntime.ToolCall{{Name: planToolName, Arguments: `{"plan":[]}`}}}}
-	if parity := empty.observePlanParity(emptyMessages); parity != taskPlanParityMatch {
-		t.Fatalf("explicit empty plan should match transcript, got %q", parity)
-	}
-}
-
-func TestTaskStateContextFallsBackWhenTranscriptDiffers(t *testing.T) {
-	state := newTaskState("objective", nil)
-	state.observe(taskStateEvent{kind: taskStateEventPlan, arguments: `{"plan":[{"content":"tracked","status":"completed"}]}`})
-	messages := []zeroruntime.Message{{Role: zeroruntime.MessageRoleAssistant, ToolCalls: []zeroruntime.ToolCall{
-		{Name: planToolName, Arguments: `{"plan":[{"content":"transcript","status":"pending"}]}`},
-	}}}
-
-	context := state.completionContext(messages, true)
-	if !context.PlanPending || context.PlanMatchesTranscript {
-		t.Fatalf("completion context must retain transcript truth on mismatch: %#v", context)
-	}
-	if context.Objective != "objective" {
-		t.Fatalf("objective lost from completion context: %#v", context)
-	}
-}
-
-func TestTaskStateCompactionSnapshotRetainsObjectiveOnPlanMismatch(t *testing.T) {
-	state := newTaskState("objective", nil)
-	state.observe(taskStateEvent{kind: taskStateEventPlan, arguments: `{"plan":[{"content":"verify","status":"pending"}]}`})
-	matching := []zeroruntime.Message{{Role: zeroruntime.MessageRoleAssistant, ToolCalls: []zeroruntime.ToolCall{
-		{Name: planToolName, Arguments: `{"plan":[{"content":"verify","status":"pending"}]}`},
-	}}}
-	if snapshot := state.snapshotForCompaction(matching); snapshot == nil || snapshot.Objective != "objective" {
-		t.Fatalf("matching transcript should produce compact state, got %#v", snapshot)
-	}
-
-	mismatching := append([]zeroruntime.Message(nil), matching...)
-	mismatching[0].ToolCalls = []zeroruntime.ToolCall{{Name: planToolName, Arguments: `{"plan":[{"content":"other","status":"pending"}]}`}}
-	if snapshot := state.snapshotForCompaction(mismatching); snapshot == nil || snapshot.Objective != "objective" || snapshot.PlanParity != taskPlanParityMismatch {
-		t.Fatalf("plan mismatch must retain immutable objective and mark mutable state uncorroborated, got %#v", snapshot)
-	}
-}
-
-func TestParseTaskPlanRejectsArgumentsTheToolWouldReject(t *testing.T) {
-	for _, arguments := range []string{
-		`{}`,
-		`{"plan":null}`,
-		`{"plan":[{"step":"alias is not accepted"}]}`,
-		`{"plan":[{"content":""}]}`,
-		`{"plan":[{"content":"valid"},{"content":""}]}`,
-		`{"plan":[{"id":4,"content":"valid"}]}`,
-	} {
-		if plan, ok := parseTaskPlan(arguments); ok {
-			t.Fatalf("parseTaskPlan(%s) = %#v, true; want rejected", arguments, plan)
-		}
-	}
-}

internal/agent/tool_result_images_test.go

@@ -1,207 +0,0 @@
-package agent
-
-import (
-	"context"
-	"fmt"
-	"strings"
-	"testing"
-
-	"github.com/Gitlawb/zero/internal/tools"
-	"github.com/Gitlawb/zero/internal/zeroruntime"
-)
-
-// imageTool returns a tool result carrying an image, the way a screenshot tool
-// should be able to.
-type imageTool struct {
-	media string
-	data  []byte
-}
-
-func (imageTool) Name() string        { return "capture" }
-func (imageTool) Description() string { return "Captures an image." }
-func (imageTool) Parameters() tools.Schema {
-	return tools.Schema{Type: "object", Properties: map[string]tools.PropertySchema{}}
-}
-func (imageTool) Safety() tools.Safety {
-	return tools.Safety{Permission: tools.PermissionAllow}
-}
-func (t imageTool) Run(context.Context, map[string]any) tools.Result {
-	return tools.Result{
-		Status: tools.StatusOK,
-		Output: "Captured a screenshot.",
-		Images: []zeroruntime.ImageBlock{{MediaType: t.media, Data: t.data}},
-	}
-}
-
-// A tool that produces an image must get that image in front of the model.
-//
-// It cannot ride the tool-result message: every provider drops images there.
-// Anthropic maps a tool result to a tool_result block with string content,
-// Gemini to a functionResponse, and OpenAI guards its image parts to the user
-// role explicitly. So the image has to arrive as a following user message,
-// which is also the only shape that keeps one tool result per tool call.
-func TestRunDeliversToolResultImagesToTheModel(t *testing.T) {
-	registry := tools.NewRegistry()
-	registry.Register(imageTool{media: "image/png", data: []byte("\x89PNG\r\n\x1a\nfake")})
-
-	provider := &mockProvider{
-		turns: [][]zeroruntime.StreamEvent{
-			{
-				{Type: zeroruntime.StreamEventToolCallStart, ToolCallID: "call_1", ToolName: "capture"},
-				{Type: zeroruntime.StreamEventToolCallDelta, ToolCallID: "call_1", ArgumentsFragment: `{}`},
-				{Type: zeroruntime.StreamEventToolCallEnd, ToolCallID: "call_1"},
-				{Type: zeroruntime.StreamEventDone},
-			},
-			{
-				{Type: zeroruntime.StreamEventText, Content: "I can see it."},
-				{Type: zeroruntime.StreamEventDone},
-			},
-		},
-	}
-
-	result, err := Run(context.Background(), "screenshot please", provider, Options{
-		Registry: registry,
-		MaxTurns: 2,
-	})
-	if err != nil {
-		t.Fatalf("Run: %v", err)
-	}
-
-	var carrier *zeroruntime.Message
-	for index := range result.Messages {
-		if len(result.Messages[index].Images) > 0 {
-			carrier = &result.Messages[index]
-			break
-		}
-	}
-	if carrier == nil {
-		t.Fatal("no message carried the tool's image; the model never saw it")
-	}
-	if carrier.Role != zeroruntime.MessageRoleUser {
-		t.Errorf("image rode a %q message; every provider only serializes images on the user role", carrier.Role)
-	}
-	if got := carrier.Images[0].MediaType; got != "image/png" {
-		t.Errorf("media type = %q, want image/png", got)
-	}
-
-	// The tool-result pairing must be untouched: one tool result per tool call.
-	toolResults := 0
-	for _, message := range result.Messages {
-		if message.Role == zeroruntime.MessageRoleTool {
-			toolResults++
-		}
-	}
-	if toolResults != 1 {
-		t.Errorf("got %d tool-result messages for 1 tool call", toolResults)
-	}
-}
-
-// textTool is a second, image-free tool so a turn can carry two calls.
-type textTool struct{}
-
-func (textTool) Name() string        { return "note" }
-func (textTool) Description() string { return "Notes something." }
-func (textTool) Parameters() tools.Schema {
-	return tools.Schema{Type: "object", Properties: map[string]tools.PropertySchema{}}
-}
-func (textTool) Safety() tools.Safety { return tools.Safety{Permission: tools.PermissionAllow} }
-func (textTool) Run(context.Context, map[string]any) tools.Result {
-	return tools.Result{Status: tools.StatusOK, Output: "noted"}
-}
-
-// Tool results for one assistant turn must stay contiguous. The image messages
-// are user-role, and a user message between two tool_results breaks strict
-// provider replay: Anthropic coalesces consecutive user content into one block
-// list and requires tool_result blocks first, so interleaving produces
-// [tool_result, text, image, tool_result] and the request is rejected.
-//
-// The first version of this feature appended each image immediately after its
-// own tool result, which is exactly that shape. A single-tool-call test passed
-// and said nothing about it.
-func TestRunKeepsToolResultsContiguousWhenAToolReturnsAnImage(t *testing.T) {
-	registry := tools.NewRegistry()
-	registry.Register(imageTool{media: "image/png", data: []byte("\x89PNG\r\n\x1a\nfake")})
-	registry.Register(textTool{})
-
-	provider := &mockProvider{turns: [][]zeroruntime.StreamEvent{
-		{
-			{Type: zeroruntime.StreamEventToolCallStart, ToolCallID: "c1", ToolName: "capture"},
-			{Type: zeroruntime.StreamEventToolCallDelta, ToolCallID: "c1", ArgumentsFragment: `{}`},
-			{Type: zeroruntime.StreamEventToolCallEnd, ToolCallID: "c1"},
-			{Type: zeroruntime.StreamEventToolCallStart, ToolCallID: "c2", ToolName: "note"},
-			{Type: zeroruntime.StreamEventToolCallDelta, ToolCallID: "c2", ArgumentsFragment: `{}`},
-			{Type: zeroruntime.StreamEventToolCallEnd, ToolCallID: "c2"},
-			{Type: zeroruntime.StreamEventDone},
-		},
-		{{Type: zeroruntime.StreamEventText, Content: "done"}, {Type: zeroruntime.StreamEventDone}},
-	}}
-
-	result, err := Run(context.Background(), "two calls", provider, Options{Registry: registry, MaxTurns: 2})
-	if err != nil {
-		t.Fatalf("Run: %v", err)
-	}
-
-	// Two separate ifs rather than a switch: a tool-role message that also
-	// carried images would match only the first arm of a switch, so a regression
-	// that put the image back on the tool result would read as "no image at all"
-	// instead of naming what actually happened.
-	type toolResult struct {
-		index int
-		id    string
-	}
-	var toolResults []toolResult
-	imageIndex := -1
-	for index, message := range result.Messages {
-		if message.Role == zeroruntime.MessageRoleTool {
-			toolResults = append(toolResults, toolResult{index: index, id: message.ToolCallID})
-		}
-		if len(message.Images) > 0 {
-			if imageIndex >= 0 {
-				t.Fatalf("image delivered twice, at %d and %d; recorded %s", imageIndex, index, messageShape(result.Messages))
-			}
-			imageIndex = index
-		}
-	}
-	if imageIndex < 0 {
-		t.Fatalf("the image never reached the model; recorded %s", messageShape(result.Messages))
-	}
-	if len(toolResults) != 2 {
-		t.Fatalf("got %d tool-result messages for 2 tool calls; recorded %s", len(toolResults), messageShape(result.Messages))
-	}
-	if toolResults[0].id != "c1" || toolResults[1].id != "c2" {
-		t.Errorf("tool results answered %q then %q, want c1 then c2", toolResults[0].id, toolResults[1].id)
-	}
-	// The contract Anthropic enforces: every tool_result block has to come before
-	// any other block in the user message it lands in. Anything wedged between
-	// the two results breaks that, which is the 400 this PR exists to avoid.
-	if toolResults[1].index != toolResults[0].index+1 {
-		t.Errorf("tool results at %d and %d are not adjacent; %q message at %d splits them; recorded %s",
-			toolResults[0].index, toolResults[1].index, result.Messages[toolResults[0].index+1].Role,
-			toolResults[0].index+1, messageShape(result.Messages))
-	}
-	// Follows the last tool result, deliberately NOT pinned to exactly last+1.
-	// appendUserBlocks coalesces consecutive user-role content, so
-	// [tool_result, tool_result, text, image] still maps to tool_result blocks
-	// first and replays fine; pinning would redden that valid shape for nothing.
-	if imageIndex <= toolResults[1].index {
-		t.Errorf("image message at %d does not follow the last tool result at %d; recorded %s",
-			imageIndex, toolResults[1].index, messageShape(result.Messages))
-	}
-}
-
-// messageShape renders the conversation as a compact role list so a failure says
-// what the ordering actually was instead of dumping whole messages.
-func messageShape(messages []zeroruntime.Message) string {
-	parts := make([]string, 0, len(messages))
-	for index, message := range messages {
-		part := fmt.Sprintf("%d:%s", index, message.Role)
-		if message.ToolCallID != "" {
-			part += "(" + message.ToolCallID + ")"
-		}
-		if len(message.Images) > 0 {
-			part += "+image"
-		}
-		parts = append(parts, part)
-	}
-	return "[" + strings.Join(parts, " ") + "]"
-}

internal/agent/turn_session.go

@@ -1,21 +0,0 @@
-package agent
-
-import (
-	"context"
-
-	"github.com/Gitlawb/zero/internal/zeroruntime"
-)
-
-// sessionProvider adapts a TurnSession back to the Provider interface so the
-// loop's existing Provider-shaped call sites (streamWithReconnect, the
-// compaction summarizer, finalAnswerAfterMaxTurns) route their stream I/O
-// through the session without any signature change. For the default adapter
-// this reduces to the wrapped provider's StreamCompletion, so behavior is
-// byte-identical; an optimized session (PR8) takes effect here transparently.
-type sessionProvider struct {
-	session zeroruntime.TurnSession
-}
-
-func (s sessionProvider) StreamCompletion(ctx context.Context, request zeroruntime.CompletionRequest) (<-chan zeroruntime.StreamEvent, error) {
-	return s.session.Stream(ctx, request)
-}

internal/agent/turn_session_test.go

@@ -1,367 +0,0 @@
-package agent
-
-import (
-	"context"
-	"errors"
-	"strings"
-	"testing"
-
-	"github.com/Gitlawb/zero/internal/tools"
-	"github.com/Gitlawb/zero/internal/zeroruntime"
-)
-
-// fakeTurnSession wraps an inner provider for streaming while counting the
-// session lifecycle, so tests can observe exactly how the loop drives it.
-type fakeTurnSession struct {
-	inner    Provider
-	prewarms int
-	streams  int
-	closes   int
-	closeErr error
-}
-
-func (s *fakeTurnSession) Prewarm(context.Context) error { s.prewarms++; return nil }
-
-func (s *fakeTurnSession) Stream(ctx context.Context, request zeroruntime.CompletionRequest) (<-chan zeroruntime.StreamEvent, error) {
-	s.streams++
-	return s.inner.StreamCompletion(ctx, request)
-}
-
-func (s *fakeTurnSession) Compact(context.Context, zeroruntime.CompletionRequest) ([]zeroruntime.Message, error) {
-	return nil, zeroruntime.ErrCompactionUnsupported
-}
-
-func (s *fakeTurnSession) Close() error { s.closes++; return s.closeErr }
-
-// fakeTurnSessionProvider opens a scripted session (or fails to).
-type fakeTurnSessionProvider struct {
-	session *fakeTurnSession
-	openErr error
-	opens   int
-}
-
-func (p *fakeTurnSessionProvider) OpenTurnSession(context.Context) (zeroruntime.TurnSession, error) {
-	p.opens++
-	if p.openErr != nil {
-		return nil, p.openErr
-	}
-	return p.session, nil
-}
-
-func (p *fakeTurnSessionProvider) Capabilities() zeroruntime.ProviderCapabilities {
-	return zeroruntime.ProviderCapabilities{}
-}
-
-func singleAnswerTurns(answer string) [][]zeroruntime.StreamEvent {
-	return [][]zeroruntime.StreamEvent{{
-		{Type: zeroruntime.StreamEventText, Content: answer},
-		{Type: zeroruntime.StreamEventDone},
-	}}
-}
-
-// TestRunOpenTurnSessionFailureSurfaces verifies a failed session open is a
-// clean run-start error: no panic, no stream ever issued, nothing to close.
-func TestRunOpenTurnSessionFailureSurfaces(t *testing.T) {
-	positional := &mockProvider{turns: singleAnswerTurns("never reached")}
-	tsp := &fakeTurnSessionProvider{openErr: errors.New("handshake refused")}
-
-	_, err := Run(context.Background(), "go", positional, Options{
-		MaxTurns:            2,
-		TurnSessionProvider: tsp,
-	})
-	if err == nil {
-		t.Fatal("expected an error from a failed session open")
-	}
-	if !strings.Contains(err.Error(), "open turn session") || !strings.Contains(err.Error(), "handshake refused") {
-		t.Fatalf("expected a wrapped open-turn-session error, got %v", err)
-	}
-	if len(positional.requests) != 0 {
-		t.Fatalf("expected no provider request after a failed open, got %d", len(positional.requests))
-	}
-	if tsp.opens != 1 {
-		t.Fatalf("expected exactly one open attempt, got %d", tsp.opens)
-	}
-}
-
-// TestRunCloseFailureIsSafe verifies a Close error at teardown is swallowed:
-// the run still returns its normal result.
-func TestRunCloseFailureIsSafe(t *testing.T) {
-	inner := &mockProvider{turns: singleAnswerTurns("done")}
-	session := &fakeTurnSession{inner: inner, closeErr: errors.New("close blew up")}
-	tsp := &fakeTurnSessionProvider{session: session}
-
-	result, err := Run(context.Background(), "go", inner, Options{
-		MaxTurns:            2,
-		TurnSessionProvider: tsp,
-	})
-	if err != nil {
-		t.Fatalf("expected Close failure to be swallowed, got %v", err)
-	}
-	if result.FinalAnswer != "done" {
-		t.Fatalf("expected normal result despite Close error, got %q", result.FinalAnswer)
-	}
-	if session.closes != 1 {
-		t.Fatalf("expected exactly one Close at teardown, got %d", session.closes)
-	}
-}
-
-// TestRunStreamsThroughTurnSession verifies that when a TurnSessionProvider is
-// wired, every model request of the run flows through the session (and the
-// session is prewarmed once and closed once).
-func TestRunStreamsThroughTurnSession(t *testing.T) {
-	inner := &mockProvider{turns: singleAnswerTurns("via session")}
-	session := &fakeTurnSession{inner: inner}
-	tsp := &fakeTurnSessionProvider{session: session}
-
-	// The positional provider is a DIFFERENT mock: it must stay untouched,
-	// proving the session (not the positional provider) carried the run.
-	positional := &mockProvider{turns: singleAnswerTurns("wrong path")}
-
-	result, err := Run(context.Background(), "go", positional, Options{
-		MaxTurns:            2,
-		TurnSessionProvider: tsp,
-	})
-	if err != nil {
-		t.Fatal(err)
-	}
-	if result.FinalAnswer != "via session" {
-		t.Fatalf("expected the session-backed answer, got %q", result.FinalAnswer)
-	}
-	if len(positional.requests) != 0 {
-		t.Fatalf("expected the positional provider to be bypassed, got %d requests", len(positional.requests))
-	}
-	if session.streams == 0 || session.streams != len(inner.requests) {
-		t.Fatalf("expected every request to flow through the session: streams=%d inner=%d", session.streams, len(inner.requests))
-	}
-	if session.prewarms != 1 {
-		t.Fatalf("expected exactly one prewarm, got %d", session.prewarms)
-	}
-	if session.closes != 1 {
-		t.Fatalf("expected exactly one close, got %d", session.closes)
-	}
-}
-
-// TestRunDefaultTurnSessionMatchesRawProvider verifies the default (nil
-// TurnSessionProvider) path is behavior-identical to passing an explicit
-// default adapter over the same provider: same result, same request count.
-func TestRunDefaultTurnSessionMatchesRawProvider(t *testing.T) {
-	run := func(explicit bool) (Result, *mockProvider) {
-		provider := &mockProvider{turns: escalateThenAnswerTurns("identical")}
-		registry := tools.NewRegistry()
-		registry.Register(escalatingTool{target: ""})
-		options := Options{Registry: registry, MaxTurns: 4}
-		if explicit {
-			options.TurnSessionProvider = zeroruntime.NewProviderTurnSessionProvider(provider, zeroruntime.ProviderCapabilities{})
-		}
-		result, err := Run(context.Background(), "go", provider, options)
-		if err != nil {
-			t.Fatal(err)
-		}
-		return result, provider
-	}
-
-	defaultResult, defaultProvider := run(false)
-	explicitResult, explicitProvider := run(true)
-
-	if defaultResult.FinalAnswer != explicitResult.FinalAnswer {
-		t.Fatalf("final answers diverged: default=%q explicit=%q", defaultResult.FinalAnswer, explicitResult.FinalAnswer)
-	}
-	if defaultResult.Turns != explicitResult.Turns {
-		t.Fatalf("turn counts diverged: default=%d explicit=%d", defaultResult.Turns, explicitResult.Turns)
-	}
-	if len(defaultProvider.requests) != len(explicitProvider.requests) {
-		t.Fatalf("request counts diverged: default=%d explicit=%d", len(defaultProvider.requests), len(explicitProvider.requests))
-	}
-}
-
-// TestRunModelSwitchClosesSessionAndContinues verifies a mid-run model switch
-// closes the active session exactly once and streams the rest of the run on
-// the new provider.
-func TestRunModelSwitchClosesSessionAndContinues(t *testing.T) {
-	registry := tools.NewRegistry()
-	registry.Register(escalatingTool{target: "claude-opus-4.1"})
-
-	firstInner := &mockProvider{turns: escalateThenAnswerTurns("unused")}
-	firstSession := &fakeTurnSession{inner: firstInner}
-	tsp := &fakeTurnSessionProvider{session: firstSession}
-
-	secondProvider := &mockProvider{turns: singleAnswerTurns("answered after switch")}
-
-	result, err := Run(context.Background(), "go", firstInner, Options{
-		Registry:            registry,
-		Model:               "claude-sonnet-4.5",
-		MaxTurns:            4,
-		TurnSessionProvider: tsp,
-		ModelSwitcher: func(_ context.Context, _ string) (Provider, error) {
-			return secondProvider, nil
-		},
-	})
-	if err != nil {
-		t.Fatal(err)
-	}
-	if result.FinalAnswer != "answered after switch" {
-		t.Fatalf("expected the post-switch answer, got %q", result.FinalAnswer)
-	}
-	// The original session handled exactly the escalation turn, then was closed
-	// once by the swap (the teardown defer closes the NEW session, not this one).
-	if firstSession.streams != 1 {
-		t.Fatalf("expected the original session to stream exactly the escalation turn, got %d", firstSession.streams)
-	}
-	if firstSession.closes != 1 {
-		t.Fatalf("expected the swap to close the original session exactly once, got %d", firstSession.closes)
-	}
-	if len(secondProvider.requests) != 1 {
-		t.Fatalf("expected the post-switch turn on the new provider, got %d requests", len(secondProvider.requests))
-	}
-}
-
-// TestRunModelSessionSwitcherCarriesOptimizedSession verifies the target-aware
-// switcher: the switched-to TurnSessionProvider's DISTINCT session receives the
-// post-switch streams, is prewarmed exactly once, and is closed exactly once at
-// teardown — and the legacy ModelSwitcher is never consulted when both are set.
-func TestRunModelSessionSwitcherCarriesOptimizedSession(t *testing.T) {
-	registry := tools.NewRegistry()
-	registry.Register(escalatingTool{target: "claude-opus-4.1"})
-
-	firstInner := &mockProvider{turns: escalateThenAnswerTurns("unused")}
-	firstSession := &fakeTurnSession{inner: firstInner}
-	initial := &fakeTurnSessionProvider{session: firstSession}
-
-	switchedInner := &mockProvider{turns: singleAnswerTurns("answered on the switched session")}
-	switchedSession := &fakeTurnSession{inner: switchedInner}
-	switched := &fakeTurnSessionProvider{session: switchedSession}
-
-	var switchedTo string
-	legacyCalled := false
-	result, err := Run(context.Background(), "go", firstInner, Options{
-		Registry:            registry,
-		Model:               "claude-sonnet-4.5",
-		MaxTurns:            4,
-		TurnSessionProvider: initial,
-		ModelSessionSwitcher: func(_ context.Context, modelID string) (zeroruntime.TurnSessionProvider, error) {
-			switchedTo = modelID
-			return switched, nil
-		},
-		ModelSwitcher: func(_ context.Context, _ string) (Provider, error) {
-			legacyCalled = true
-			return nil, errors.New("legacy switcher must not be consulted")
-		},
-	})
-	if err != nil {
-		t.Fatal(err)
-	}
-	if result.FinalAnswer != "answered on the switched session" {
-		t.Fatalf("expected the switched session's answer, got %q", result.FinalAnswer)
-	}
-	if switchedTo != "claude-opus-4.1" {
-		t.Fatalf("expected switch to claude-opus-4.1, got %q", switchedTo)
-	}
-	if legacyCalled {
-		t.Fatal("legacy ModelSwitcher was consulted despite ModelSessionSwitcher being set")
-	}
-	if switched.opens != 1 {
-		t.Fatalf("expected exactly one open of the switched provider, got %d", switched.opens)
-	}
-	// The switched session carries the rest of the run: it streams the answer
-	// turn, is prewarmed exactly once at the swap, and is closed exactly once by
-	// the teardown defer.
-	if switchedSession.streams != 1 || len(switchedInner.requests) != 1 {
-		t.Fatalf("expected the switched session to stream the post-switch turn, streams=%d inner=%d", switchedSession.streams, len(switchedInner.requests))
-	}
-	if switchedSession.prewarms != 1 {
-		t.Fatalf("expected the switched session to be prewarmed exactly once, got %d", switchedSession.prewarms)
-	}
-	if switchedSession.closes != 1 {
-		t.Fatalf("expected the switched session to be closed exactly once at teardown, got %d", switchedSession.closes)
-	}
-	// The original session ends with the swap: one stream, one close.
-	if firstSession.streams != 1 || firstSession.closes != 1 {
-		t.Fatalf("expected the original session to end at the swap (streams=%d closes=%d)", firstSession.streams, firstSession.closes)
-	}
-}
-
-// TestRunModelSessionSwitcherOpenFailureIsNonFatal verifies the swap-time open
-// failure: the switcher returns a TurnSessionProvider whose OpenTurnSession
-// fails. The run stays non-fatal on the ORIGINAL session (closed only at
-// teardown), and the switch-failure note reaches the next request.
-func TestRunModelSessionSwitcherOpenFailureIsNonFatal(t *testing.T) {
-	registry := tools.NewRegistry()
-	registry.Register(escalatingTool{target: "claude-opus-4.1"})
-
-	inner := &mockProvider{turns: escalateThenAnswerTurns("recovered after failed open")}
-	session := &fakeTurnSession{inner: inner}
-	failingOpen := &fakeTurnSessionProvider{openErr: errors.New("switched handshake refused")}
-
-	result, err := Run(context.Background(), "go", inner, Options{
-		Registry:            registry,
-		Model:               "claude-sonnet-4.5",
-		MaxTurns:            4,
-		TurnSessionProvider: &fakeTurnSessionProvider{session: session},
-		ModelSessionSwitcher: func(_ context.Context, _ string) (zeroruntime.TurnSessionProvider, error) {
-			return failingOpen, nil
-		},
-	})
-	if err != nil {
-		t.Fatalf("expected a failed swap open to be non-fatal, got %v", err)
-	}
-	if result.FinalAnswer != "recovered after failed open" {
-		t.Fatalf("expected the run to recover on the original session, got %q", result.FinalAnswer)
-	}
-	if failingOpen.opens != 1 {
-		t.Fatalf("expected exactly one open attempt on the switched provider, got %d", failingOpen.opens)
-	}
-	// The original session serves both turns and is closed only at teardown —
-	// never by the failed swap.
-	if session.streams != 2 || session.closes != 1 {
-		t.Fatalf("expected the original session to serve both turns and close once (streams=%d closes=%d)", session.streams, session.closes)
-	}
-	var sawNote bool
-	for _, message := range inner.requests[1].Messages {
-		if message.Role == zeroruntime.MessageRoleUser && strings.Contains(strings.ToLower(message.Content), "could not switch") {
-			sawNote = true
-		}
-	}
-	if !sawNote {
-		t.Fatalf("expected a switch-failure note on the next turn, messages: %+v", inner.requests[1].Messages)
-	}
-}
-
-// TestRunModelSessionSwitcherErrorIsNonFatal verifies a ModelSessionSwitcher
-// error keeps the run on the current session with a transcript note — the same
-// non-fatal contract as ModelSwitcher.
-func TestRunModelSessionSwitcherErrorIsNonFatal(t *testing.T) {
-	registry := tools.NewRegistry()
-	registry.Register(escalatingTool{target: "claude-opus-4.1"})
-
-	inner := &mockProvider{turns: escalateThenAnswerTurns("recovered on original session")}
-	session := &fakeTurnSession{inner: inner}
-
-	result, err := Run(context.Background(), "go", inner, Options{
-		Registry:            registry,
-		Model:               "claude-sonnet-4.5",
-		MaxTurns:            4,
-		TurnSessionProvider: &fakeTurnSessionProvider{session: session},
-		ModelSessionSwitcher: func(_ context.Context, _ string) (zeroruntime.TurnSessionProvider, error) {
-			return nil, errors.New("session build blew up")
-		},
-	})
-	if err != nil {
-		t.Fatalf("expected session-switcher error to be non-fatal, got %v", err)
-	}
-	if result.FinalAnswer != "recovered on original session" {
-		t.Fatalf("expected the run to recover on the original session, got %q", result.FinalAnswer)
-	}
-	// Both turns stream through the original session; it closes once at teardown.
-	if session.streams != 2 || session.closes != 1 {
-		t.Fatalf("expected the original session to serve both turns and close once (streams=%d closes=%d)", session.streams, session.closes)
-	}
-	var sawNote bool
-	for _, request := range inner.requests[1].Messages {
-		if request.Role == zeroruntime.MessageRoleUser && strings.Contains(strings.ToLower(request.Content), "could not switch") {
-			sawNote = true
-		}
-	}
-	if !sawNote {
-		t.Fatalf("expected a non-fatal switch-failure note on the next turn, messages: %+v", inner.requests[1].Messages)
-	}
-}

internal/agent/types.go

@@ -3,12 +3,10 @@ package agent
 import (
 	"context"
 
-	"github.com/Gitlawb/zero/internal/execution"
 	"github.com/Gitlawb/zero/internal/hooks"
 	"github.com/Gitlawb/zero/internal/sandbox"
 	"github.com/Gitlawb/zero/internal/streamjson"
 	"github.com/Gitlawb/zero/internal/tools"
-	"github.com/Gitlawb/zero/internal/trace"
 	"github.com/Gitlawb/zero/internal/zeroruntime"
 )
 
@@ -26,15 +24,6 @@ const (
 	PermissionModeAsk       PermissionMode = "ask"
 	PermissionModeUnsafe    PermissionMode = "unsafe"
 	PermissionModeSpecDraft PermissionMode = "spec-draft"
-	// PermissionModePlan is an interactive, read-only planning mode. It applies
-	// to the CURRENT session (unlike spec-draft, which drafts in a separate
-	// session): the agent may inspect the workspace and shape the plan with
-	// update_plan/ask_user, but no mutating tool is advertised, so it cannot
-	// write files, run shell, or implement while planning. Entry points:
-	// the TUI's /plan on (exit with /plan off, which restores whatever mode
-	// was active before), `zero exec --plan`, and the ACP session mode
-	// selector ("plan").
-	PermissionModePlan PermissionMode = "plan"
 	// PermissionModeMemberAuto is a headless mode for swarm/specialist MEMBERS: it
 	// advertises the in-workspace mutators a member needs to build (write/edit +
 	// shell) on top of the Auto set, while the sandbox engine still gates them at
@@ -71,33 +60,17 @@ const (
 )
 
 type ToolResult struct {
-	ToolCallID string
-	Name       string
-	Status     tools.Status
-	Output     string
-	// Truncated reports that the tool's model-visible output omitted content.
-	// The full result may be recoverable through Meta["spill_path"].
-	Truncated bool
-	Meta      map[string]string
-	// Images the tool produced, delivered to the model as a following user
-	// message rather than on this result. See tools.Result.Images.
-	Images       []zeroruntime.ImageBlock
+	ToolCallID   string
+	Name         string
+	Status       tools.Status
+	Output       string
+	Meta         map[string]string
 	Redacted     bool
 	ChangedFiles []string
-	// ChangeSummaries are non-selectable generated-tree summaries emitted by
-	// command execution; callers must not schedule per-file work from them.
-	ChangeSummaries []execution.Change
-	Display         tools.Display
-	Outcome         tools.ToolOutcome
+	Display      tools.Display
 	// DenialReason categorizes why a tool call was blocked (empty when it ran).
 	// It lets a surface distinguish the cause precisely instead of parsing Output.
 	DenialReason DenialCategory
-	// Risk is the sandbox risk classification of this call, stamped for EXECUTED
-	// results so run-policy observers (the execution-profile controller) can see
-	// the risk level of an allowed mutation. It mirrors the classification the
-	// permission path already computes; denied or canceled results keep the zero
-	// value. Pure observation: nothing about permissions or sandboxing changes.
-	Risk sandbox.Risk
 	// LoadedTools carries the deferred-tool names a tool_search call asked the
 	// loop to expose next turn (lifted from Meta["load_tools"]). nil for every
 	// ordinary tool result; only tool_search populates it.
@@ -109,24 +82,6 @@ type ToolResult struct {
 	RequestedModel string
 }
 
-// ModelOutput returns the bounded provider-facing result while preserving
-// compatibility with synthetic and restored results created before outcomes
-// were finalized.
-func (result ToolResult) ModelOutput() string {
-	if result.Outcome.Finalized() {
-		return result.Outcome.ModelView
-	}
-	return result.Output
-}
-
-// HumanDisplay returns the presentation intended for interactive surfaces.
-func (result ToolResult) HumanDisplay() tools.Display {
-	if result.Outcome.Finalized() {
-		return result.Outcome.HumanView
-	}
-	return result.Display
-}
-
 // DenialCategory classifies why a tool call was blocked before it executed.
 type DenialCategory string
 
@@ -139,54 +94,6 @@ const (
 	DenialHookBlocked      DenialCategory = "hook_blocked"      // vetoed by a beforeTool hook
 )
 
-// ProfilePolicy is the loop-facing slice of a selected execution profile.
-// The surface (exec flag, TUI command) resolves a named profile into this
-// policy; the loop only ever sees the policy, never the catalog.
-type ProfilePolicy struct {
-	// Name labels trace counters and diagnostics (e.g. "fast"). The trace
-	// recorder's profile label is set by the caller at recorder construction.
-	Name string
-	// Escalate, when non-nil, arms one-shot in-run posture escalation.
-	Escalate *PostureEscalation
-}
-
-// PostureEscalation describes a one-shot escalation to stricter knob values,
-// applied mid-run when an armed trigger fires. Targets are the values the
-// selected profile DISPLACED at run start (i.e. "restore the balanced
-// posture"), so escalation can never introduce a value that was not already
-// valid for this run and model. Zero-valued targets leave that knob untouched.
-type PostureEscalation struct {
-	// MaxTurns raises the turn ceiling to this value when greater than the
-	// ceiling in effect. 0 leaves the ceiling untouched.
-	MaxTurns int
-	// ReasoningEffort replaces the run's effort when non-empty.
-	ReasoningEffort string
-	// RestoreDefaultEffort clears the run's effort override (back to the
-	// provider/model default) when true. It exists because the displaced value
-	// of a profile-filled effort is "" — which as a ReasoningEffort target
-	// means "leave untouched" — so restoring the default needs its own signal.
-	// Ignored when ReasoningEffort is non-empty.
-	RestoreDefaultEffort bool
-	// RestoreCompletionGate re-enables RequireCompletionSignal (headless
-	// completion semantics) when true.
-	RestoreCompletionGate bool
-
-	// Triggers. A zero value disables that signal entirely.
-	// OnToolFailureStreak fires when the repeated-failure guard observes a
-	// same-tool retriable-failure streak of at least this length.
-	OnToolFailureStreak int
-	// OnCompletionUncertain fires on the Nth uncertain completion evaluation
-	// (continue nudge or semantic check). Headless only: the completion gate
-	// never runs interactively.
-	OnCompletionUncertain int
-	// OnSelfCorrectFailure fires when a post-edit verification cycle reports a
-	// failing outcome (correcting, reported, or aborted).
-	OnSelfCorrectFailure bool
-	// OnRiskyMutation fires when an EXECUTED tool result carries a sandbox risk
-	// level at or above this threshold. Empty disables the signal.
-	OnRiskyMutation sandbox.RiskLevel
-}
-
 type PermissionRequest struct {
 	ToolCallID         string                     `json:"toolCallId"`
 	ToolName           string                     `json:"name"`
@@ -204,22 +111,6 @@ type PermissionRequest struct {
 	Grant              *sandbox.Grant             `json:"grant,omitempty"`
 	CommandPrefix      []string                   `json:"commandPrefix,omitempty"`
 	AvailableDecisions []PermissionDecisionAction `json:"availableDecisions,omitempty"`
-	// PrefixApprovalEscalates reports that approving a command prefix will also
-	// run the command OUTSIDE the sandbox, not merely stop asking about it.
-	//
-	// It exists because that consequence was real but invisible. Approving a
-	// prefix rewrites the call to sandbox_permissions: require_escalated, which
-	// resolves to a nil engine and genuinely unsandboxed execution, and the
-	// engine's own escalation prompt is then satisfied by the approval just
-	// given for the sandboxed form. So the operator authorized one thing and got
-	// a wider one, having been shown only "allow command prefix for session".
-	//
-	// The escalation itself is deliberate and guarded: proposedCommandPrefix
-	// refuses to offer a prefix while any other segment of the command is not
-	// known-safe, precisely so an unreviewed segment cannot ride out of the
-	// sandbox on it. What was missing was telling the person deciding, so this
-	// surfaces it rather than removing it.
-	PrefixApprovalEscalates bool `json:"prefixApprovalEscalates,omitempty"`
 }
 
 type PermissionDecision struct {
@@ -325,12 +216,8 @@ type Options struct {
 	ProviderName     string
 	Model            string
 	ReasoningEffort  string
-	ServiceTier      string
 	Cwd              string
 	SystemPrompt     string
-	// TransientSystemPrompt adds trusted runtime guidance for this run only.
-	// Empty preserves the ordinary system prompt byte-for-byte.
-	TransientSystemPrompt string
 	// ResponseStyle is the operator-selected reply style from the TUI /style
 	// command (e.g. "concise", "explanatory", "review"). It is rendered into the
 	// system prompt as a short directive. Empty or "balanced" adds nothing — the
@@ -381,11 +268,9 @@ type Options struct {
 	// specialist child process emits while running. The toolCallID identifies
 	// which Task tool call the progress belongs to. nil is a no-op.
 	OnToolProgress func(toolCallID string, event streamjson.Event)
-	// OnContext, when set, is called for each main agent request with its
-	// per-category context budget, including a replacement request after
-	// compaction or a stall retry. Internal summarizer requests are excluded so
-	// surfaces keep showing the active conversation budget. Opt-in like the other
-	// callbacks; nil is a no-op.
+	// OnContext, when set, is called once per turn with the per-category context
+	// budget of the request about to be sent, so a surface (TUI/CLI) can show
+	// context utilization. Opt-in like the other callbacks; nil is a no-op.
 	OnContext func(ContextBreakdown)
 	// ModelSwitcher, when set, lets a tool escalate the run to a stronger model
 	// mid-run: the loop calls it with the requested model id and, on success,
@@ -394,37 +279,6 @@ type Options struct {
 	// request), so every existing caller is unaffected. A returned error is
 	// non-fatal: the run continues on the current model.
 	ModelSwitcher func(ctx context.Context, modelID string) (Provider, error)
-	// TurnSessionProvider, when set, supplies the turn session the run streams
-	// through — the seam an optimized provider session (connection reuse,
-	// prewarm, native compaction) plugs into without touching the loop. nil
-	// keeps the default: the loop wraps the passed provider in a no-op session
-	// whose Stream IS provider.StreamCompletion, so behavior is byte-identical
-	// and every existing caller is unaffected.
-	TurnSessionProvider zeroruntime.TurnSessionProvider
-	// ModelSessionSwitcher, when set, is the target-aware escalation hook: the
-	// loop prefers it over ModelSwitcher, and its TurnSessionProvider keeps an
-	// optimized session (and its capabilities) across a mid-run model switch.
-	// nil falls back to ModelSwitcher, whose bare Provider is wrapped in the
-	// default no-op session — today's behavior, unchanged. Same non-fatal error
-	// contract as ModelSwitcher: a returned error records a note and the run
-	// continues on the current model and session.
-	ModelSessionSwitcher func(ctx context.Context, modelID string) (zeroruntime.TurnSessionProvider, error)
-	// Profile, when set, arms the execution-profile posture controller for this
-	// run (auto-escalation to stricter knob values on failure/uncertainty/risky
-	// mutation signals). nil — the default everywhere today — leaves the loop
-	// byte-identical: no observation, no escalation, no counters. Same opt-in
-	// convention as Trace and SelfCorrect.
-	Profile *ProfilePolicy
-	// Trace, when set, records per-turn timing for the run: the loop stamps
-	// spans (prompt build, generation, tool execution, permission wait,
-	// compaction, provider connect) and counters (model requests, tool calls,
-	// retries, tokens) into it. nil DISABLES tracing entirely — every stamp is
-	// nil-safe and the loop is byte-identical to an untraced run. The caller
-	// owns the recorder: Run stamps into it but does not Finish or emit it.
-	// A fresh Recorder is required per Run — reusing one across runs merges
-	// their spans, counters, and first-event timestamps, and Finish freezes a
-	// recorder so no further stamps take.
-	Trace *trace.Recorder
 	// SelfCorrect, when set, runs a post-edit verify-and-correct cycle after a
 	// mutating tool call: it verifies the changed files (LSP diagnostics + project
 	// tests) and feeds failures back to the model to fix, bounded by an attempt
@@ -447,11 +301,8 @@ type Options struct {
 	// continuation cue ("…Let me check the config:"). The loop then nudges the
 	// model to continue instead, bounded by maxContinueNudges (and still by
 	// MaxTurns and the run deadline); if the model keeps stalling, the run
-	// finalizes as INCOMPLETE (Result.Incomplete) rather than success. When the
-	// run's profile also enables SelfCorrect, an otherwise-complete turn gets one
-	// task-grounded semantic check before success; profiles without SelfCorrect
-	// add no model call. Default false leaves the loop byte-identical, so the
-	// interactive TUI is unaffected.
+	// finalizes as INCOMPLETE (Result.Incomplete) rather than success. Default
+	// false leaves the loop byte-identical, so the interactive TUI is unaffected.
 	RequireCompletionSignal bool
 
 	runPermissions *permissionRunState
@@ -480,6 +331,13 @@ type Result struct {
 	IncompleteReason string
 }
 
+// Truncated reports whether the final response ended abnormally (cut off at the
+// output token cap or withheld by a content filter) rather than completing
+// naturally. Callers can use it to warn the user that FinalAnswer is incomplete.
+func (result Result) Truncated() bool {
+	return result.FinishReason != ""
+}
+
 // TruncationNotice returns a user-facing warning when the final response was
 // truncated, or "" for a normal completion. Shared by the CLI and TUI so the
 // wording stays consistent.

internal/agenteval/agent_command.go

@@ -31,6 +31,10 @@ type AgentRunner interface {
 
 type AgentRunnerFunc func(context.Context, AgentRunInput) AgentRunResult
 
+func (fn AgentRunnerFunc) Run(ctx context.Context, input AgentRunInput) AgentRunResult {
+	return fn(ctx, input)
+}
+
 // defaultAgentOutputLimit caps captured stdout/stderr per stream so a chatty or
 // runaway agent cannot exhaust memory or bloat the benchmark report.
 const defaultAgentOutputLimit = 1 << 20 // 1 MiB per stream

internal/agenteval/export_test.go

@@ -1,8 +0,0 @@
-// Test seams: helpers only test code uses, kept out of the production binary.
-package agenteval
-
-import "context"
-
-func (fn AgentRunnerFunc) Run(ctx context.Context, input AgentRunInput) AgentRunResult {
-	return fn(ctx, input)
-}

internal/agenteval/score.go

@@ -125,7 +125,9 @@ func Score(suite Suite, input ScoreInput) Report {
 	if len(task.RequiredTraceEvents) > 0 {
 		report.Results = append(report.Results, scoreTraceEvents(task.RequiredTraceEvents, input))
 	}
-	report.Results = append(report.Results, unknownCommandResults(input.CommandResults, seenCommands, input)...)
+	for _, result := range unknownCommandResults(input.CommandResults, seenCommands, input) {
+		report.Results = append(report.Results, result)
+	}
 	report.finishSummary()
 	return report
 }

internal/agentinit/agentinit.go

@@ -40,14 +40,14 @@ func FormatFacts(info repoinfo.Info) string {
 	b.WriteString("Pre-computed repository facts (from a local scan — verify and fill gaps):\n")
 
 	if info.PrimaryLanguage != "" {
-		fmt.Fprintf(&b, "- Primary language: %s (of %d detected)\n", info.PrimaryLanguage, info.LanguageCount)
+		b.WriteString(fmt.Sprintf("- Primary language: %s (of %d detected)\n", info.PrimaryLanguage, info.LanguageCount))
 	}
 	if langs := topLanguages(info.Languages, 5); langs != "" {
 		b.WriteString("- Languages: " + langs + "\n")
 	}
-	fmt.Fprintf(&b, "- Size: ~%d files, ~%d LOC, max dir depth %d\n", info.FileCount, info.LOCEstimate, info.MaxDepth)
+	b.WriteString(fmt.Sprintf("- Size: ~%d files, ~%d LOC, max dir depth %d\n", info.FileCount, info.LOCEstimate, info.MaxDepth))
 	if info.WorkspaceType != "" && info.WorkspaceType != "none" {
-		fmt.Fprintf(&b, "- Workspace: %s (%d packages)\n", info.WorkspaceType, info.WorkspacePackageCount)
+		b.WriteString(fmt.Sprintf("- Workspace: %s (%d packages)\n", info.WorkspaceType, info.WorkspacePackageCount))
 	}
 	if len(info.BuildTools) > 0 {
 		b.WriteString("- Build tools: " + strings.Join(info.BuildTools, ", ") + "\n")

internal/aimlapi/client.go

@@ -1,259 +0,0 @@
-package aimlapi
-
-import (
-	"bytes"
-	"context"
-	"encoding/json"
-	"fmt"
-	"io"
-	"net/http"
-	"net/url"
-	"strings"
-	"time"
-)
-
-// PaymentMethod selects how a partner-checkout top-up is paid for.
-type PaymentMethod string
-
-const (
-	PaymentMethodCard   PaymentMethod = "card"
-	PaymentMethodCrypto PaymentMethod = "crypto"
-)
-
-// SessionStatus is the lifecycle state of a partner-checkout session, as reported
-// by the aimlapi.com backend while polling for payment.
-type SessionStatus string
-
-const (
-	SessionStatusPendingAuth    SessionStatus = "pending_auth"
-	SessionStatusPendingPayment SessionStatus = "pending_payment"
-	SessionStatusPaid           SessionStatus = "paid"
-	SessionStatusExchanging     SessionStatus = "exchanging"
-	SessionStatusExchanged      SessionStatus = "exchanged"
-	SessionStatusCancelled      SessionStatus = "cancelled"
-	SessionStatusExpired        SessionStatus = "expired"
-	SessionStatusFailed         SessionStatus = "failed"
-)
-
-// AuthResult is the bearer token minted by a passwordless sign-in / sign-up.
-type AuthResult struct {
-	Token string `json:"token"`
-	Exp   int64  `json:"exp"`
-}
-
-// PartnerCheckoutSession is a partner-attributed top-up session; its SessionToken
-// addresses it on later pay/exchange/poll calls.
-type PartnerCheckoutSession struct {
-	ID             string        `json:"id"`
-	SessionToken   string        `json:"sessionToken"`
-	PartnerID      string        `json:"partnerId"`
-	PartnerName    *string       `json:"partnerName"`
-	UserID         *int64        `json:"userId"`
-	AmountUSDMinor *int64        `json:"amountUsdMinor"`
-	Status         SessionStatus `json:"status"`
-	IssuedKeyID    *string       `json:"issuedKeyId"`
-	ReturnURL      *string       `json:"returnUrl"`
-}
-
-// PaymentSession is the hosted payment-provider checkout; PayURL is the page the
-// user opens to pay.
-type PaymentSession struct {
-	ProviderSessionID string `json:"providerSessionId"`
-	PayURL            string `json:"payUrl"`
-}
-
-// PayResult pairs the hosted checkout with the updated partner-checkout session
-// returned when a top-up payment is initiated.
-type PayResult struct {
-	Checkout        PaymentSession         `json:"checkout"`
-	PartnerCheckout PartnerCheckoutSession `json:"partnerCheckout"`
-}
-
-// ExchangeResult holds the API key minted when a paid session is exchanged.
-type ExchangeResult struct {
-	APIKey   string `json:"apiKey"`
-	APIKeyID string `json:"apiKeyId"`
-}
-
-// TopUpByKeyRequest funds a partner-checkout session with the API key that owns
-// the account (Path A / env key), instead of an email session.
-type TopUpByKeyRequest struct {
-	SessionToken     string
-	AmountUSDMinor   int
-	PaymentSessionID string // idempotency: same id → same checkout, never a second charge
-	AutoTopUp        bool   // enroll the account in automatic top-up at checkout (card, saved off-session)
-	SuccessURL       string
-	CancelURL        string
-}
-
-// TopUpByKeyResult is the hosted checkout returned by TopUpByKey.
-type TopUpByKeyResult struct {
-	Checkout PaymentSession `json:"checkout"`
-}
-
-// APIError is a non-2xx HTTP response from the aimlapi.com API. Status carries the
-// code so callers can branch (e.g. 401 = bad key, 5xx = retryable).
-type APIError struct {
-	Message string
-	Status  int
-	Body    string
-}
-
-// Error renders the method, endpoint, status, and (when present) the response body.
-func (e APIError) Error() string {
-	if strings.TrimSpace(e.Body) != "" {
-		return fmt.Sprintf("%s: HTTP %d: %s", e.Message, e.Status, strings.TrimSpace(e.Body))
-	}
-	return fmt.Sprintf("%s: HTTP %d", e.Message, e.Status)
-}
-
-// Client talks to the aimlapi.com auth, app, and inference APIs for the onboarding
-// and partner-checkout top-up flows.
-type Client struct {
-	endpoints  Endpoints
-	httpClient *http.Client
-}
-
-func NewClient(endpoints Endpoints, httpClient *http.Client) *Client {
-	if httpClient == nil {
-		httpClient = &http.Client{Timeout: 60 * time.Second}
-	}
-	return &Client{endpoints: endpoints, httpClient: httpClient}
-}
-
-// CreateSession opens a partner-checkout session attributed to partnerID; the
-// returned SessionToken addresses it on the pay/exchange/poll calls.
-func (c *Client) CreateSession(ctx context.Context, partnerID string, partnerName string, returnURL string) (PartnerCheckoutSession, error) {
-	body := map[string]any{"partnerId": partnerID}
-	if strings.TrimSpace(partnerName) != "" {
-		body["partnerName"] = strings.TrimSpace(partnerName)
-	}
-	if strings.TrimSpace(returnURL) != "" {
-		body["returnUrl"] = strings.TrimSpace(returnURL)
-	}
-	var result PartnerCheckoutSession
-	err := c.request(ctx, http.MethodPost, strings.TrimRight(c.endpoints.AppBaseURL, "/")+"/v3/partner-checkout/sessions", "", body, &result)
-	return result, err
-}
-
-// GetSession fetches the current state of a partner-checkout session; used to poll
-// for payment completion.
-func (c *Client) GetSession(ctx context.Context, sessionToken string) (PartnerCheckoutSession, error) {
-	var result PartnerCheckoutSession
-	err := c.request(ctx, http.MethodGet, strings.TrimRight(c.endpoints.AppBaseURL, "/")+"/v3/partner-checkout/sessions/"+urlPathEscape(sessionToken), "", nil, &result)
-	return result, err
-}
-
-// Pay initiates payment for a session and returns the hosted checkout URL.
-// autoTopUp enrolls the account in automatic top-up at checkout time.
-func (c *Client) Pay(ctx context.Context, bearer string, sessionToken string, amountUSDMinor int, paymentSessionID string, method PaymentMethod, successURL string, cancelURL string, autoTopUp bool) (PayResult, error) {
-	body := map[string]any{
-		"amountUsdMinor":   amountUSDMinor,
-		"paymentSessionId": paymentSessionID,
-		"method":           method,
-	}
-	if strings.TrimSpace(successURL) != "" {
-		body["successUrl"] = strings.TrimSpace(successURL)
-	}
-	if strings.TrimSpace(cancelURL) != "" {
-		body["cancelUrl"] = strings.TrimSpace(cancelURL)
-	}
-	// Only sent when enabled: enrolls the account in automatic top-up. The backend
-	// honours this field at checkout time.
-	if autoTopUp {
-		body["autoTopUp"] = true
-	}
-	var result PayResult
-	err := c.request(ctx, http.MethodPost, strings.TrimRight(c.endpoints.AppBaseURL, "/")+"/v3/partner-checkout/sessions/"+urlPathEscape(sessionToken)+"/pay", bearer, body, &result)
-	return result, err
-}
-
-// Exchange converts a paid session into a freshly minted API key (new-account flow).
-func (c *Client) Exchange(ctx context.Context, bearer string, sessionToken string) (ExchangeResult, error) {
-	var result ExchangeResult
-	err := c.request(ctx, http.MethodPost, strings.TrimRight(c.endpoints.AppBaseURL, "/")+"/v3/partner-checkout/sessions/"+urlPathEscape(sessionToken)+"/exchange", bearer, nil, &result)
-	return result, err
-}
-
-// TopUpByKey funds a partner-checkout session using the raw API key that owns the
-// account (Path A / env key), returning the hosted checkout. It binds the session
-// to the key's account server-side, so no email session is needed and no key is
-// exchanged. req.PaymentSessionID makes a retry idempotent (same id → same
-// checkout, never a second charge).
-func (c *Client) TopUpByKey(ctx context.Context, apiKey string, req TopUpByKeyRequest) (TopUpByKeyResult, error) {
-	body := map[string]any{
-		"sessionToken":     req.SessionToken,
-		"amountUsdMinor":   req.AmountUSDMinor,
-		"paymentSessionId": req.PaymentSessionID,
-	}
-	if req.AutoTopUp {
-		body["autoTopUp"] = true
-	}
-	if strings.TrimSpace(req.SuccessURL) != "" {
-		body["successUrl"] = strings.TrimSpace(req.SuccessURL)
-	}
-	if strings.TrimSpace(req.CancelURL) != "" {
-		body["cancelUrl"] = strings.TrimSpace(req.CancelURL)
-	}
-	var result TopUpByKeyResult
-	err := c.request(ctx, http.MethodPost, c.billingV2URL("/billing/topup"), apiKey, body, &result)
-	return result, err
-}
-
-// billingV2URL builds a v2 billing API URL from the inference base. The top-up
-// endpoint lives next to the balance endpoint on the API gateway, one version up
-// (".../v1" → ".../v2/billing/..."), so it re-anchors the version while keeping any
-// host/prefix an override supplied.
-func (c *Client) billingV2URL(path string) string {
-	base := strings.TrimRight(strings.TrimSpace(c.endpoints.InferenceBaseURL), "/")
-	base = strings.TrimSuffix(base, "/v1")
-	return strings.TrimRight(base, "/") + "/v2" + path
-}
-
-func (c *Client) request(ctx context.Context, method string, endpoint string, bearer string, body any, out any) error {
-	var reader io.Reader
-	if body != nil {
-		data, err := json.Marshal(body)
-		if err != nil {
-			return err
-		}
-		reader = bytes.NewReader(data)
-	}
-	request, err := http.NewRequestWithContext(ctx, method, endpoint, reader)
-	if err != nil {
-		return err
-	}
-	request.Header.Set("Accept", "application/json")
-	if body != nil {
-		request.Header.Set("Content-Type", "application/json")
-	}
-	if strings.TrimSpace(bearer) != "" {
-		request.Header.Set("Authorization", "Bearer "+strings.TrimSpace(bearer))
-	}
-	response, err := c.httpClient.Do(request)
-	if err != nil {
-		return fmt.Errorf("network request to %s failed: %w", endpoint, err)
-	}
-	defer response.Body.Close()
-	text, readErr := io.ReadAll(io.LimitReader(response.Body, 1<<20))
-	if readErr != nil {
-		return readErr
-	}
-	if response.StatusCode < 200 || response.StatusCode >= 300 {
-		return APIError{Message: method + " " + endpoint, Status: response.StatusCode, Body: string(text)}
-	}
-	if out == nil {
-		return nil
-	}
-	if len(bytes.TrimSpace(text)) == 0 {
-		return APIError{Message: method + " " + endpoint + " returned empty body", Status: response.StatusCode}
-	}
-	if err := json.Unmarshal(text, out); err != nil {
-		return APIError{Message: method + " " + endpoint + " returned non-JSON body", Status: response.StatusCode, Body: string(text)}
-	}
-	return nil
-}
-
-func urlPathEscape(value string) string {
-	return url.PathEscape(value)
-}

internal/aimlapi/client_onboard.go

@@ -1,106 +0,0 @@
-package aimlapi
-
-import (
-	"context"
-	"fmt"
-	"net/http"
-	"strings"
-)
-
-// Additional client methods backing the interactive TUI onboarding flow. They
-// target the endpoints shipped in the aimlapi backend:
-//   - GET  {inference}/billing/balance            (Path A: validate key + balance)
-//   - PATCH{auth}/v1/auth/account                 (Path B: does the account exist?)
-//   - POST {auth}/v1/auth/sign-in/code(/verify)   (Path B existing: passwordless login)
-//   - POST {auth}/v1/auth/account/passwordless    (Path B new: email-only signup)
-//   - POST {app}/v1/keys                          (mint a key once we hold a session)
-
-// BalanceResult is the wallet balance returned by GetBalance, with the backend's
-// low-balance flag and threshold.
-type BalanceResult struct {
-	Balance             float64 `json:"balance"`
-	LowBalance          bool    `json:"lowBalance"`
-	LowBalanceThreshold float64 `json:"lowBalanceThreshold"`
-}
-
-// CheckResult reports whether an email already has an account, driving the
-// sign-in vs sign-up branch of the onboarding flow.
-type CheckResult struct {
-	// "sign-in" (account exists) or "sign-up" (no account yet).
-	Action   string  `json:"action"`
-	Provider *string `json:"provider"`
-}
-
-// CreatedKey is an API key minted by CreateKey.
-type CreatedKey struct {
-	Key string `json:"key"`
-	ID  string `json:"id"`
-}
-
-func (c *Client) authURL(path string) string {
-	return strings.TrimRight(c.endpoints.AuthBaseURL, "/") + path
-}
-
-func (c *Client) appURL(path string) string {
-	return strings.TrimRight(c.endpoints.AppBaseURL, "/") + path
-}
-
-func (c *Client) inferenceURL(path string) string {
-	return strings.TrimRight(c.endpoints.InferenceBaseURL, "/") + path
-}
-
-// GetBalance doubles as key validation: a 401 means the key is invalid, a 2xx
-// returns the wallet balance. Auth is the raw API key itself.
-func (c *Client) GetBalance(ctx context.Context, apiKey string) (BalanceResult, error) {
-	var result BalanceResult
-	err := c.request(ctx, http.MethodGet, c.inferenceURL("/billing/balance"), apiKey, nil, &result)
-	return result, err
-}
-
-// CheckAccount reports whether an email already has an aimlapi.com account
-// (Action "sign-in") or not ("sign-up").
-func (c *Client) CheckAccount(ctx context.Context, email string) (CheckResult, error) {
-	var result CheckResult
-	err := c.request(ctx, http.MethodPatch, c.authURL("/v1/auth/account"), "", map[string]any{"email": email}, &result)
-	return result, err
-}
-
-// SendSignInCode emails a 6-digit sign-in code to an existing account.
-func (c *Client) SendSignInCode(ctx context.Context, email string) error {
-	return c.request(ctx, http.MethodPost, c.authURL("/v1/auth/sign-in/code"), "", map[string]any{"email": email}, nil)
-}
-
-// VerifySignInCode exchanges an emailed code for a session bearer token.
-func (c *Client) VerifySignInCode(ctx context.Context, email string, code string) (AuthResult, error) {
-	var result AuthResult
-	err := c.request(ctx, http.MethodPost, c.authURL("/v1/auth/sign-in/code/verify"), "", map[string]any{"email": email, "code": code}, &result)
-	if err == nil && strings.TrimSpace(result.Token) == "" {
-		err = fmt.Errorf("aimlapi.com did not return an auth token")
-	}
-	return result, err
-}
-
-// CreatePasswordlessAccount registers a new email-only account and returns its
-// session bearer token.
-func (c *Client) CreatePasswordlessAccount(ctx context.Context, email string) (AuthResult, error) {
-	var result AuthResult
-	err := c.request(ctx, http.MethodPost, c.authURL("/v1/auth/account/passwordless"), "", map[string]any{"email": email}, &result)
-	if err == nil && strings.TrimSpace(result.Token) == "" {
-		err = fmt.Errorf("aimlapi.com did not return an auth token")
-	}
-	return result, err
-}
-
-// CreateKey mints a fresh API key for a session-holding user.
-func (c *Client) CreateKey(ctx context.Context, bearer string, name string) (CreatedKey, error) {
-	body := map[string]any{}
-	if strings.TrimSpace(name) != "" {
-		body["name"] = strings.TrimSpace(name)
-	}
-	var result CreatedKey
-	err := c.request(ctx, http.MethodPost, c.appURL("/v1/keys"), bearer, body, &result)
-	if err == nil && strings.TrimSpace(result.Key) == "" {
-		err = fmt.Errorf("aimlapi.com did not return an API key")
-	}
-	return result, err
-}

internal/aimlapi/client_test.go

@@ -1,159 +0,0 @@
-package aimlapi
-
-import (
-	"context"
-	"encoding/json"
-	"errors"
-	"net/http"
-	"net/http/httptest"
-	"testing"
-)
-
-func TestCreateKeySendsAuthenticatedRequestAndDecodesKey(t *testing.T) {
-	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
-		if r.Method != http.MethodPost || r.URL.Path != "/v1/keys" {
-			t.Fatalf("request = %s %s, want POST /v1/keys", r.Method, r.URL.Path)
-		}
-		if got := r.Header.Get("Authorization"); got != "Bearer session-token" {
-			t.Fatalf("Authorization = %q", got)
-		}
-		var body map[string]string
-		if err := json.NewDecoder(r.Body).Decode(&body); err != nil {
-			t.Fatal(err)
-		}
-		if body["name"] != "Zero CLI" {
-			t.Fatalf("name = %q", body["name"])
-		}
-		w.Header().Set("Content-Type", "application/json")
-		_, _ = w.Write([]byte(`{"key":"sk-issued","id":"key-id"}`))
-	}))
-	defer server.Close()
-
-	client := NewClient(Endpoints{AppBaseURL: server.URL}, server.Client())
-	key, err := client.CreateKey(context.Background(), " session-token ", " Zero CLI ")
-	if err != nil {
-		t.Fatal(err)
-	}
-	if key.Key != "sk-issued" || key.ID != "key-id" {
-		t.Fatalf("key = %#v", key)
-	}
-}
-
-func TestCreateKeyRejectsEmptyKeyResponse(t *testing.T) {
-	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, _ *http.Request) {
-		w.Header().Set("Content-Type", "application/json")
-		_, _ = w.Write([]byte(`{"id":"key-id"}`))
-	}))
-	defer server.Close()
-
-	client := NewClient(Endpoints{AppBaseURL: server.URL}, server.Client())
-	if _, err := client.CreateKey(context.Background(), "session-token", "Zero"); err == nil {
-		t.Fatal("CreateKey accepted a response without an API key")
-	}
-}
-
-func TestAuthMethodsRejectEmptyTokenResponses(t *testing.T) {
-	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, _ *http.Request) {
-		w.Header().Set("Content-Type", "application/json")
-		_, _ = w.Write([]byte(`{"token":""}`))
-	}))
-	defer server.Close()
-
-	client := NewClient(Endpoints{AuthBaseURL: server.URL}, server.Client())
-	tests := []struct {
-		name string
-		call func() error
-	}{
-		{
-			name: "verify sign-in code",
-			call: func() error {
-				_, err := client.VerifySignInCode(context.Background(), "[email protected]", "123456")
-				return err
-			},
-		},
-		{
-			name: "create passwordless account",
-			call: func() error {
-				_, err := client.CreatePasswordlessAccount(context.Background(), "[email protected]")
-				return err
-			},
-		},
-	}
-	for _, test := range tests {
-		t.Run(test.name, func(t *testing.T) {
-			if err := test.call(); err == nil {
-				t.Fatal("accepted a response without an auth token")
-			}
-		})
-	}
-}
-
-func TestClientReturnsTypedAPIError(t *testing.T) {
-	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, _ *http.Request) {
-		http.Error(w, "invalid key", http.StatusUnauthorized)
-	}))
-	defer server.Close()
-
-	client := NewClient(Endpoints{InferenceBaseURL: server.URL}, server.Client())
-	_, err := client.GetBalance(context.Background(), "bad-key")
-	var apiErr APIError
-	if !errors.As(err, &apiErr) {
-		t.Fatalf("error = %T %v, want APIError", err, err)
-	}
-	if apiErr.Status != http.StatusUnauthorized || apiErr.Body != "invalid key\n" {
-		t.Fatalf("APIError = %#v", apiErr)
-	}
-}
-
-func TestTypedResponseRejectsEmptySuccessBody(t *testing.T) {
-	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, _ *http.Request) {
-		w.WriteHeader(http.StatusNoContent)
-	}))
-	defer server.Close()
-
-	client := NewClient(Endpoints{InferenceBaseURL: server.URL}, server.Client())
-	if _, err := client.GetBalance(context.Background(), "key"); err == nil {
-		t.Fatal("GetBalance accepted an empty successful response")
-	}
-}
-
-func TestBodylessOperationAcceptsEmptySuccessBody(t *testing.T) {
-	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, _ *http.Request) {
-		w.WriteHeader(http.StatusNoContent)
-	}))
-	defer server.Close()
-
-	client := NewClient(Endpoints{AuthBaseURL: server.URL}, server.Client())
-	if err := client.SendSignInCode(context.Background(), "[email protected]"); err != nil {
-		t.Fatalf("SendSignInCode() error = %v", err)
-	}
-}
-
-func TestParseAmountUSD(t *testing.T) {
-	tests := []struct {
-		name    string
-		value   string
-		want    int
-		wantErr bool
-	}{
-		{name: "default", want: DefaultAmountUSDMinor},
-		{name: "minimum", value: "20", want: 2000},
-		{name: "cents", value: "25.49", want: 2549},
-		{name: "below minimum", value: "19.99", wantErr: true},
-		{name: "above maximum", value: "10001", wantErr: true},
-		{name: "not a number", value: "twenty", wantErr: true},
-		{name: "nan", value: "NaN", wantErr: true},
-		{name: "infinity", value: "Inf", wantErr: true},
-	}
-	for _, test := range tests {
-		t.Run(test.name, func(t *testing.T) {
-			got, err := ParseAmountUSD(test.value)
-			if (err != nil) != test.wantErr {
-				t.Fatalf("ParseAmountUSD(%q) error = %v", test.value, err)
-			}
-			if got != test.want {
-				t.Fatalf("ParseAmountUSD(%q) = %d, want %d", test.value, got, test.want)
-			}
-		})
-	}
-}