CedarCopilot wants to merge 1 commit into main from staging
Not deployed
Libra has no production signal for this change yet because it has not deployed. Libra checks hourly for 3 days after each deploy.
This PR updates task completion and reopening behavior, draft cleanup timing, document persistence handling, agenda synchronization, and sidebar conversation typing.
Two reliability gaps remain: reopening an agent-enabled task can be reported as successful even when its execution was not scheduled, and leaving the page during the completion Undo window can leave a Gmail draft behind.
Not safe to merge until reopened agent tasks cannot be confirmed as reopened after their required scheduling fails. The draft-cleanup gap is non-blocking but leaves stale Gmail drafts that require manual cleanup.
A reproduced task-execution failure materially affects the reliability of reopened agent tasks. A separate reproduced cleanup gap affects draft lifecycle behavior without blocking the core completion action.
Files Needing Attention: apps/server/src/trpc/routes/user-tasks.ts needs failure handling or durable retry for reopening schedules. apps/mail/modules/userTasks/hooks/use-optimistic-task-actions.ts needs durable delayed draft cleanup.
What T-Rex did
apps/server/src/trpc/routes/user-tasks.ts, line 1337-1351 (link)
Report reopen scheduling failures
Reopening an agent-enabled task writes its status as todo, then logs and ignores a failed scheduling attempt before returning success. When scheduling infrastructure rejects the task before creating its queue or scheduling record, the user sees a successful reopen but the task will not receive its expected agent execution. Return the scheduling failure, or persist a retryable scheduling request, rather than confirming the reopen as successful. This must be fixed before merging.
Knowledge Base Used:
Libra has not measured any production surfaces for this change yet.