Skip to content

needs-triage survives a claim: claimed (and closed) self-reports still read "Triage needed" on the activity monitor #14644

Closed

Description

@clubanderson

Symptom

Owner, 2026-10-27 17:40Z, on the Hive activity monitor: "wth does triage-needed mean on a claimed issue?"

Live example: #14586 [selfreport] agent_pane_timeout — claimed by the scanner at 17:36Z (stage/020-claimed), closed at 17:40Z, and still wearing needs-triage (GraphQL timeline: needs-triage was added at filing, 15:25:40Z, and has no UnlabeledEvent at all). The tracker row read Claimed with a "triage needed" chip; the chip only "disappeared" because the row dropped off when the issue closed.

Root cause

  1. src/pkg/fleetreport/runtime.go:682 (SelfReportLabels) stamps needs-triage on every hive-filed self-report at filing.
  2. src/pkg/github/issue_stage_labels.go:188 (issueUntriaged) derives stage/015-triage-needed purely from that label; 015 is documented as a tracking state, not a gate (stagelabels.go:96-105, transition_table.go allows 015→020), so the scanner claims it like a 010.
  3. stagelabels.Plan only ever touches hive-owned labels, so the 015→020 write left needs-triage on. The only remover (reporter_trust_wait.go, marker-gated) never fires for fleetreport-filed issues.
  4. src/pkg/dashboard/stage_tracker.go:~1170 deliberately kept the needs-triage modifier on non-015 stages, so the row read "triage needed" on a Claimed issue.

Net effect: every hive-filed self-report that gets claimed carries needs-triage forever, and the activity monitor reads it as a live signal.

Fix

  • Stage engine: when the derived stage settles triage (020, 030, 040 or a PR-mirrored stage), Decision.Clear names needs-triage and stagelabels.Plan removes it in the same label write as the stage (direct 015→020 claim included). The label survives at 010/017/045, which the transition table lets fall back to 015.
  • Sweep: an open issue already at 020+ still wearing needs-triage is reconciled on the next pass (one write, logged through the existing stage-writer outcome).
  • Dashboard: stage text is derived from the stage label only; a needs-triage left on a 020+ row is dropped from modifiers and reported as stale_labels with a muted Stale label tooltip line. No "Accept triage" action on a claimed row.
  • Docs (labels-and-control-signals.md, dashboard-sections.md) and changelog.

Related observations (not fixed here)

  • convergence shadow: candidate would be WITHHELD … blockers=["needs-triage"] (src/cmd/hive/convergence_kick.go) — enforcing this for needs-triage would withhold hive-filed self-reports from the scanner forever (nobody else triages them). Recommend not enforcing it for needs-triage; clearing the label (this fix) is the right lever. Whether fleetreport should stamp needs-triage on hive-filed reports at all is a separate question.
  • The shadow fence would have denied external mutation … writers exhausted (max 1) warnings seen on [selfreport] agent_pane_timeout: prompt timeout — dumping pane #14586 are a separate repo-wide mutation-boundary capacity issue — filed separately.

Activity

  1. added
    help wantedDenotes an issue that needs help from a contributor. Must meet "help wanted" guidelines.
    on Oct 27, 2026
  2. hivecommons-hive commented on Oct 27, 2026

    @hivecommons-hive
    Contributor

    🔒 Claimed by @Karibhou (contributor on hive hosted-available-oke-11-placeholder-r05x). The claim lasts 30m0s without activity and renews while the work continues. Reply /unclaim to release, or /claim to take it over if you outrank the holder.

  3. Karibhou commented on Oct 27, 2026

    @Karibhou
    Contributor

    Checked before starting work: this issue is already fully covered by open PR #14648 (base v5, not a draft, body carries Fixes #14644). It implements every item in the Fix section: stagelabels Desired.Clear + Stage.SettlesTriage(), stageengine Decision.Clear on 020/030/040/PR-mirrored stages, the sweep reconcile for already-claimed issues, the dashboard stale_labels tooltip with no "Accept triage" on claimed rows, docs, changelog and tests. I'm not opening a duplicate PR; this issue will close when #14648 merges.

    — hive: backend=claude model=claude-opus-5-5


    🐝 Hive Agent: contributor | SHA: 68f8dd297

  4. added a commit that references this issue on Oct 27, 2026
    2cd109b
  5. hivecommons-hive commented on Oct 27, 2026

    @hivecommons-hive
    Contributor

    🔓 Claim by @Karibhou released (lease revoked).

  6. github-actions commented on Oct 27, 2026

    @github-actions
    Contributor

    @clubanderson this issue was closed by @hivecommons-hive[bot] as completed.

    If this was closed by a merged PR, the fix only takes effect once it ships in a tagged release and your deployment updates to it — check your running version against the fix before concluding the fix failed.

    If you are still seeing the problem, comment /reopen and it will be reopened for you — you do not need write access. Please add what you are still observing (running version, logs, steps) so it is not closed the same way again.

  7. added a commit that references this issue on Oct 27, 2026
    8e9c592
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    help wantedDenotes an issue that needs help from a contributor. Must meet "help wanted" guidelines.kind/bugCategorizes issue or PR as related to a bug.

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions