fix(subagents): report startup and completion lifecycle edges#77415
fix(subagents): report startup and completion lifecycle edges#77415neilofneils404 wants to merge 4 commits into
Conversation
|
Thanks for the context here. I swept through the related work, and this is now duplicate or superseded. Close as duplicate/superseded: the remaining startup lifecycle work is already tracked by #58776, while this PR is no longer a safe landing branch because it is conflicting and still targets retired Root-cause cluster Members:
Proposal only: this assessment does not dispatch repair, suppress jobs, mutate sibling items, close, or merge anything. Canonical path: Close this stale branch and keep #58776 as the canonical work item for a fresh current-main implementation of the lifecycle contract. So I’m closing this here and keeping the remaining discussion on #58776. Review detailsBest possible solution: Close this stale branch and keep #58776 as the canonical work item for a fresh current-main implementation of the lifecycle contract. Do we have a high-confidence way to reproduce the issue? Yes, at source level. Current main lacks the requested phases, active consumers still handle only existing lifecycle states, and aborted terminal events still flow through timeout/killed paths instead of a typed startup-failure signal. Is this the best way to solve the issue? No, not as submitted. The lifecycle direction is plausible, but this stale PR targets retired files and should be replaced by a current-main implementation through the active embedded-agent, terminal outcome, gateway, session, SDK, and subagent surfaces. Security review: Security review cleared: No concrete security or supply-chain concern was found; the diff changes in-repo lifecycle logic, tests, docs, and changelog text only. AGENTS.md: found and applied where relevant. What I checked:
Likely related people:
Codex review notes: model internal, reasoning high; reviewed against 5361e5a0b455. |
98ed21d to
4e27a08
Compare
(cherry picked from commit d6624ba17c7301b5784b730ddd80030720091fab)
(cherry picked from commit 2969df1) (cherry picked from commit 77d4777f9975e1db692ad35a5ed181907495ba8f) (cherry picked from commit 73dd5a227be8e63968755e25b162c067dd611e0e)
dcdff46 to
83aefea
Compare
|
This pull request has been automatically marked as stale due to inactivity. |
Summary
Validation
pnpm check:no-conflict-markersnode scripts/run-vitest.mjs run --config test/vitest/vitest.agents.config.ts src/agents/subagent-announce.test.ts src/agents/subagent-registry-lifecycle.test.tsnode scripts/run-vitest.mjs run --config test/vitest/vitest.agents-pi-embedded.config.ts src/agents/pi-embedded-subscribe.handlers.lifecycle.test.ts src/agents/pi-embedded-subscribe.handlers.tools.test.ts src/agents/subagent-registry.test.tsReal behavior proof
Behavior or issue addressed: Gateway agent waits now preserve a
startup-failedphase even when a trailing completion/end edge arrives, so callers can observe the startup failure instead of it being overwritten by a generic end state.Real environment tested: Local OpenClaw checkout on SpicyTaco (macOS/Darwin), PR branch
cass/subagent-lifecycle-fixes-20260504at83aefea574after rebase ontoorigin/maind38e30e02c.Exact steps or command run after this patch:
node scripts/run-vitest.mjs run --config test/vitest/vitest.gateway.config.ts src/gateway/server-methods/server-methods.test.tsEvidence after fix: Terminal capture from the local OpenClaw checkout after the fix:
Observed result after fix: The focused gateway method regression suite passes on the patched branch, including the startup-failed lifecycle regression coverage added for this behavior.
What was not tested: I did not force a live production Gateway subagent startup failure; the after-fix behavior was verified in the local OpenClaw gateway method harness with terminal output above.