 John Smith
|
f7d450a1d9
|
docs(agents): record the multi-agent studio invariants, harnesses and current fork divergence
|
2026-09-24 10:01:09 -04:00 |
|
|
John Smith
|
84cb3299e0
|
docs(autonomous): completion audit against the objective
|
2026-09-24 05:52:54 -04:00 |
|
|
John Smith
|
22349d89c9
|
feat(ai): steer a sub-agent that stops making progress
|
2026-09-24 05:52:30 -04:00 |
|
|
John Smith
|
a9ab2861d8
|
feat(ai): fallback to a secondary model when the provider keeps failing
|
2026-09-24 05:45:55 -04:00 |
|
|
John Smith
|
1eb4896ac3
|
docs(autonomous): adversarial-invariant audit and full editor-suite gate
|
2026-09-24 05:39:49 -04:00 |
|
|
John Smith
|
cbba33b651
|
fix(ai): retry a transient provider 502/503 instead of failing the turn
|
2026-09-24 05:37:42 -04:00 |
|
|
John Smith
|
600d472d29
|
docs(autonomous): verify the manager-role fix end-to-end
|
2026-09-24 05:33:00 -04:00 |
|
|
John Smith
|
f05870eb1a
|
fix(ai): run the top-level orchestrator as the studio manager role
|
2026-09-24 05:25:39 -04:00 |
|
|
John Smith
|
d79549c534
|
fix(ai): advertise the array/object args the change tools read
|
2026-09-24 05:16:48 -04:00 |
|
|
John Smith
|
5b675fa904
|
fix(ai): align change_behavior_property schema with its implementation
|
2026-09-24 05:12:10 -04:00 |
|
|
John Smith
|
77968102f5
|
fix(ai): keep a live local request processing after it loses selection
|
2026-09-24 05:08:53 -04:00 |
|
|
John Smith
|
11ed7dad8f
|
docs(autonomous): run @feedback-loop end-to-end against the hosted proxy
|
2026-09-24 04:56:20 -04:00 |
|
|
John Smith
|
236d85a1e8
|
docs(autonomous): real-model end-to-end run of the multi-agent flow
|
2026-09-24 04:40:57 -04:00 |
|
|
John Smith
|
853e6dcc98
|
feat(ai): show each sub-agent's transcript inline in the dashboard
|
2026-09-24 04:34:08 -04:00 |
|
|
John Smith
|
a11eb64de3
|
feat(ai): link spawn calls to their plan task in the chat
|
2026-09-24 04:28:21 -04:00 |
|
|
John Smith
|
5a4f147172
|
feat(ai): resume a stalled studio manager with a bounded nudge
|
2026-09-24 04:19:19 -04:00 |
|
|
John Smith
|
ec740df324
|
feat(ai): reveal each finished sub-agent's report inline in the dashboard
|
2026-09-24 04:02:20 -04:00 |
|
|
John Smith
|
dc6f7d5a2c
|
feat(ai): persist per-agent token totals across reloads
|
2026-09-24 03:58:15 -04:00 |
|
|
John Smith
|
fb88efdf33
|
feat(ai): add an expandable per-agent audit dashboard to the chat
|
2026-09-24 03:53:31 -04:00 |
|
|
John Smith
|
6403010c4e
|
chore(autonomous): re-run the real-model benchmark with the session's fixes
|
2026-09-24 03:48:41 -04:00 |
|
|
John Smith
|
4c51fa1475
|
chore(autonomous): whole-surface lint and format verification
|
2026-09-24 03:44:08 -04:00 |
|
|
John Smith
|
81d9b2fe60
|
docs(autonomous): assess loop-guard thresholds and close the coverage sweep
|
2026-09-24 03:43:28 -04:00 |
|
|
John Smith
|
1642f0dab0
|
test(ai): pin the backend-only tool allowlist
|
2026-09-24 03:42:44 -04:00 |
|
|
John Smith
|
a11239869f
|
chore(autonomous): full-suite regression and run-log refresh
|
2026-09-24 03:41:30 -04:00 |
|
|
John Smith
|
c7a503fa23
|
fix(ai): deliver all batch outcomes to the model on mixed batches
|
2026-09-24 03:40:59 -04:00 |
|
|
John Smith
|
6e5ccc3793
|
fix(ai): deliver the loop-guard warning to the model
|
2026-09-24 03:39:27 -04:00 |
|
|
John Smith
|
fa44919225
|
fix(ai): send role-denied calls to the model instead of looping
|
2026-09-24 03:35:40 -04:00 |
|
|
John Smith
|
fa8e02cdf1
|
chore(autonomous): full-suite regression and run-log refresh
|
2026-09-24 03:29:18 -04:00 |
|
|
John Smith
|
aba290662d
|
docs(autonomous): record the per-role specialist flow in the run log
|
2026-09-24 03:28:47 -04:00 |
|
|
John Smith
|
36e0c728b5
|
chore(scripts): commit the reproducible specialist-flow script
|
2026-09-24 03:28:41 -04:00 |
|
|
John Smith
|
7df2ac6efd
|
chore(autonomous): deterministic per-role specialist flow with a project open
|
2026-09-24 03:28:19 -04:00 |
|
|
John Smith
|
b609cef480
|
feat(scripts): script each sub-agent role separately in the mock
|
2026-09-24 03:26:20 -04:00 |
|
|
John Smith
|
e679114f11
|
fix(ai): reuse the null-tolerant finalization check in AiRequestContext
|
2026-09-24 03:25:29 -04:00 |
|
|
John Smith
|
9ec579d2d2
|
fix(ai): make the audit summary's role map prototype-safe
|
2026-09-24 03:23:02 -04:00 |
|
|
John Smith
|
5aa3a1b4d9
|
refactor(ai): extract the per-agent token sum to the tested leaf
|
2026-09-24 03:21:45 -04:00 |
|
|
John Smith
|
493d55f681
|
chore(autonomous): build and deterministic regression after the retry change
|
2026-09-24 03:18:45 -04:00 |
|
|
John Smith
|
552f237727
|
feat(ai): retry once when the endpoint is unreachable
|
2026-09-24 03:15:48 -04:00 |
|
|
John Smith
|
12f0f11942
|
chore(autonomous): full-suite regression and run-log refresh
|
2026-09-24 03:13:12 -04:00 |
|
|
John Smith
|
affbb64212
|
docs(autonomous): record the streaming-counter live-catch attempt
|
2026-09-24 03:12:33 -04:00 |
|
|
John Smith
|
bc2e1ad69d
|
feat(ai): show a live streaming token count in the chat
|
2026-09-24 03:11:31 -04:00 |
|
|
John Smith
|
039c82d7ca
|
docs(autonomous): verify Flow over the session additions
|
2026-09-24 03:07:39 -04:00 |
|
|
John Smith
|
26ceba5347
|
docs(autonomous): assess an upstream-file guard and revert on policy
|
2026-09-24 03:07:11 -04:00 |
|
|
John Smith
|
735314d00b
|
docs(autonomous): re-verify the no-paywall invariant
|
2026-09-24 03:05:42 -04:00 |
|
|
John Smith
|
681ef85548
|
chore(autonomous): full-suite regression and run-log refresh
|
2026-09-24 03:05:11 -04:00 |
|
|
John Smith
|
22ada3a276
|
fix(ai): keep sub-agent children out of persisted chats
|
2026-09-24 03:04:37 -04:00 |
|
|
John Smith
|
3bef2bf1c5
|
fix(ai): clear a request's provider telemetry when deleted
|
2026-09-24 03:02:52 -04:00 |
|
|
John Smith
|
04cc35dce5
|
feat(ai): bill sub-agent tokens to the sub-agent and show them
|
2026-09-24 03:01:51 -04:00 |
|
|
John Smith
|
32aaea1dba
|
chore(autonomous): full-suite regression and run-log refresh
|
2026-09-24 02:58:10 -04:00 |
|
|
John Smith
|
8e1056deb4
|
feat(ai): break the multi-agent audit summary down by role
|
2026-09-24 02:56:04 -04:00 |
|
|
John Smith
|
4a1e93c0ed
|
fix(ai): count distinct sub-agents in the audit summary
|
2026-09-24 02:53:13 -04:00 |
|