PR monitor chips never retire — survive PR close and merge, session archive, and app restart; no way to clear them
Claude Desktop's GitHub-integration PR monitors don't auto-archive when their PR closes — or even when it merges (a MERGED PR with green checks still shows a live monitor chip). The chips also survive archiving the session that created them and full app restarts, and there's no programmatic surface (MCP tool / CLI) to clear them. Meanwhile the monitor's "Auto-fix CI & address comments" automation stays armed against PRs whose head branches have been deleted (chip shows "No checks reported yet").
Repro:
- Open several PRs from a session (gh CLI).
- Land them locally (rebase + ff-merge + push) — GitHub marks some Closed rather than Merged depending on ref-push order; merge one normally so it's MERGED.
- Delete the head branches; archive the originating session; restart the app.
- All monitor chips persist indefinitely — Closed, Merged, green or gray alike.
Requests:
- Auto-archive monitors on PR close and merge by default (or expose that in Auto-archive settings).
- A per-chip "stop monitoring / archive now" action.
- Disarm "Auto-fix CI" when the PR is closed or its head branch is deleted.
Environment: macOS 26.5 (arm64), Claude Desktop 1.9659.2 with bundled Claude Code; GitHub integration connected.
Related (PR-monitoring family, not duplicates): #62928 (monitoring no-ops on some CI setups), #63222 (status-polling rate limits), #64460 (wake-message config) — this issue covers monitor lifecycle/retirement, which none of those address.
3 Comments
Found 3 possible duplicate issues:
This issue will be automatically closed as a duplicate in 3 days.
🤖 Generated with Claude Code
Closing as duplicate of #66055 (chips never clear; branch-name keying — corroborated there with our cross-repo same-branch-name case) and #61247 (no unpin/remove affordance in the CI monitoring panel). Unique datapoints posted on #66055.
This issue has been automatically locked since it was closed and has not had any activity for 7 days. If you're experiencing a similar issue, please file a new issue and reference this one if it's relevant.