You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
🐛 Red+stale claim release drops a covered issue to Actionable while the hive is still fixing its PR #16522
#16475 is covered by open PR #16492 (Closes #16475, hive-authored, base v5). The PR went red on one check and the reviewer asked for changes, so the PR sits at stage/070-fixing and the governor reaper re-dispatched a fix for it (reaper: re-dispatched fix for stuck red PR … re_engagements=1).
Ten minutes after the red head stopped moving, the Fix #3 claim-release valve (FilterClaimedIssues → claimingPRRedStale) reported the claiming PR red+stale and released the issue: releasing issue: claiming PR red+stale. The issue-stage sweep then wrote stage/040-covered-by-pr → stage/010-actionable (the Activity Monitor row shows Actionable with a stage ↓ regression pill) and the scanner kick lists the issue as plain actionable work with no mention of the open PR.
Why it is wrong
Two parts of the hive now contradict each other about the same work:
The PR lanes say 070 Fixing: the hive owns the repair and is spending a re-engagement budget on it (MaxReEngagements spaced by ReEngageCooldown, so the fix loop runs for ≥ 1 h).
The issue lanes say 010 Actionable: nobody owns this, take it.
The valve predates the re-engagement budget and the stage engine. Releasing the issue at the 10-minute mark, while the reaper is still working the PR, invites a second PR for the same issue. Today the live open-PR dispatch guard (OpenPRLinkGuard, #14812) refuses the scanner's claim and labels the issue has-open-pr, so the duplicate is usually prevented — but the guard fails open when the PR listing errors, and every cycle still produces a wrong stage label, a misleading stage ↓ on the dashboard, and a kick offer the scanner cannot accept.
Expected
An issue whose covering PR is red+stale stays stage/040-covered-by-pr (reason narrating the PR's 070 stage) for as long as the hive is still re-engaging that PR. The claim is released only once the hive has given up on the PR: it was escalated to a human (needs-human), the re-engagement budget for the red head is exhausted, or the red head has sat untouched for the whole re-engagement window (MaxReEngagements × ReEngageCooldown) without a single delivered re-engagement (unreachable owner).
@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.
Activity
github-actions commentedon Nov 4, 2026
@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
/reopenand 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.Merge remote-tracking branch 'origin/v5' into v6-topup-20261103 (top-…