Commit Graph

8522 Commits

Author SHA1 Message Date
ZiyaZhang
666ef95f81 fix(tool-goal): let the model deliver a wrap-up message after a goal-round complete/blocked
A goal round reporting complete or blocked used to conclude the turn at
the tool result, so the model never spoke after the call and sessions
ended on a bare update_goal card. The terminal update now defers one
plugin-sourced <goal_complete>/<goal_blocked> instruction onto its
result asking for a grounded closing message without further tool
calls; the turn then ends through the ordinary no-tool-calls stop.
Direct-human mutations stay uninstructed. Wording chosen by A/B
sampling on deepseek-v4-pro; one extra request per goal lifecycle.

New keyless ACP snapshot goal-wrapup drives the shipped app through
create -> round one -> autonomous complete and pins the injection, the
same-turn closing message, and the completed turn end.
2026-08-02 05:58:45 -07:00
ZiyaZhang
04d43e3b19 feat(llm-replay): resolve {{fromRequest:...}} placeholders against the live request
A scripted sidecar cannot know values minted at run time, so terminal
goal updates (which must echo the random goal id) were previously
un-scriptable. Placeholders in scripted entries now resolve against the
request corpus at stream time: last match wins, capture group 1 or the
whole match substitutes, and unmatched/invalid/unterminated patterns
fail loud.
2026-08-02 05:58:31 -07:00
Tianyi Cui
baf133d134 Merge pull request #933 from deepseek-harness/codex/web-steer-support
feat(web): steer queued messages into active turns
2026-08-02 18:00:12 +08:00
imccyu
7d08e43720 fix(web): address steering review feedback 2026-08-02 17:53:07 +08:00
imccyu
cf918fa05d test(web): explain steering replay pace 2026-08-02 17:53:07 +08:00
imccyu
d59ab5596e docs(host): align queue mirror contract 2026-08-02 17:53:07 +08:00
imccyu
a31b3eb7b8 test(host): cover queue mirror event expiry 2026-08-02 17:53:07 +08:00
imccyu
53531b9af2 test(snapshot): refresh Cordis steering API fixture 2026-08-02 17:53:07 +08:00
imccyu
3a9f78c55e feat(web): restore actions after steering consumption 2026-08-02 17:53:07 +08:00
imccyu
dffe955ed2 feat(web): surface and configure composer steering 2026-08-02 17:53:07 +08:00
kingwl
5b17d3bab5 fix(web): match queue steer icon to design 2026-08-02 17:53:07 +08:00
kingwl
955a12cca4 feat(web): steer queued messages into active turns 2026-08-02 17:53:07 +08:00
Ziya
fff0172208 Merge pull request #1125 from deepseek-harness/agent/web-search-timeout-budget
fix(cli): give shipped web search a 60s budget
2026-08-02 05:42:41 -04:00
Tianyi Cui
089f921663 Merge pull request #983 from deepseek-harness/codex/install-pnpm-stderr
fix(install): show pnpm version errors
2026-08-02 17:41:43 +08:00
Wenlu Wang
677f7bc530 Merge pull request #1147 from deepseek-harness/codex/todo-goal-queue-layout
fix(web): align composer context stack
2026-08-02 14:53:26 +08:00
kingwl
cf70a52dcc Merge origin/master into codex/todo-goal-queue-layout 2026-08-02 14:34:13 +08:00
kingwl
6332b5b952 fix(web): align context cards on narrow screens 2026-08-02 13:26:08 +08:00
imccyu
ab320ba991 Merge pull request #1114 from deepseek-harness/codex/gui-subagents-navigation
fix(web): preserve nested subagent navigation
2026-08-02 13:03:32 +08:00
imccyu
23680e838b fix(web): synchronize subagent navigation state 2026-08-02 12:51:11 +08:00
imccyu
b94d2f9c1d fix(web): stabilize nested subagent navigation 2026-08-02 12:51:11 +08:00
imccyu
53ae9abea2 style(web): polish subagent and bash chrome 2026-08-02 12:51:11 +08:00
imccyu
ca1d838afc style(web): refine subagent navigation chrome 2026-08-02 12:51:11 +08:00
imccyu
42d34473af fix(web): allow follow-ups to running subagents 2026-08-02 12:51:11 +08:00
imccyu
130410bb98 fix(web): preserve subagent navigation and fork grouping 2026-08-02 12:51:11 +08:00
imccyu
8c9cd4c15d feat: optimize subagent list children query 2026-08-02 12:51:11 +08:00
imccyu
5d56019d22 feat: revert crumbs navigation for subagents 2026-08-02 12:51:11 +08:00
Dudu-0223
431fb4b035 feat(subagent): add explicit child reports 2026-08-02 12:51:10 +08:00
imccyu
a1b3bebb61 fix(subagent): scope drains and soften final flush 2026-08-02 12:51:10 +08:00
imccyu
d7153768f5 test(web): cover post-fork subagent continuation 2026-08-02 12:51:10 +08:00
Dudu-0223
9a7be21b7f fix(host): preserve subagent continuation ownership 2026-08-02 12:51:10 +08:00
imccyu
1aedb23ca6 test(web): align subagent access mode golden 2026-08-02 12:51:09 +08:00
imccyu
6a02570e8f test(web): close rebased subagent coverage gaps 2026-08-02 12:51:09 +08:00
imccyu
5ef5feb01a fix(web): reconcile rebased subagent contracts 2026-08-02 12:51:09 +08:00
imccyu
d950150cdc fix(persistence): preserve subagent session origin 2026-08-02 12:51:09 +08:00
imccyu
9d77a50c23 test(cli): include list_agents in shipped catalog 2026-08-02 12:51:09 +08:00
imccyu
b8be3583e0 test(cordis): refresh session origin API snapshot 2026-08-02 12:51:09 +08:00
imccyu
5113b831e8 test(client): align card fixtures with subagent session state 2026-08-02 12:51:09 +08:00
Dudu-0223
8a518e353b feat(web): rewrite subagent conversations for FIFO activation 2026-08-02 12:51:09 +08:00
Dudu-0223
f0ab04273d fix(web): deduplicate subagent navigation 2026-08-02 12:51:09 +08:00
Dudu-0223
16ffd63115 feat(web): add nested subagent conversations 2026-08-02 12:51:09 +08:00
Dudu-0223
a27492507d feat(web): add subagent conversation transport 2026-08-02 12:51:09 +08:00
imccyu
9adfc576da chore(knip): include subagent settlement fixture 2026-08-02 12:51:09 +08:00
Dudu-0223
a977ef30ee fix(subagent): preserve published run failures 2026-08-02 12:51:09 +08:00
Dudu-0223
8b0a7a5d8d fix(subagent): close stacked catalog integration gaps 2026-08-02 12:51:09 +08:00
Dudu-0223
a967f5174a test(snapshot): update unlabeled workflow children 2026-08-02 12:51:09 +08:00
Dudu-0223
03d95d0d84 fix(subagent): keep one-shot labels optional 2026-08-02 12:51:09 +08:00
Dudu-0223
774ee34b9a feat(subagent): catalog one-shot child sessions 2026-08-02 12:51:09 +08:00
Dudu-0223
de6e572e51 fix(subagent): make session query peer optional 2026-08-02 12:51:09 +08:00
Dudu-0223
2e5c439ecf fix(subagent): decouple child listing runtime 2026-08-02 12:51:09 +08:00
Dudu-0223
30b9d0464a fix: address codex review round 3
Recheck the listing cancellation signal after each child inspection settles:
a per-child read failing with a diagnostic-mapped code during an abort
previously bypassed the inspection's own checkpoints, letting a cancelled
single-child scan return a successful result.
2026-08-02 12:51:09 +08:00