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
Dashboard badges never-worked issues as 'regressed' (lapsed claim 020→010, 030↔020 flicker, partial-merge 040→010) and the marker never clears on forward moves #14570
On the hosted hive dashboard (2026-10-27 ~14:00Z) the work-queue / stage tracker shows a red regressed chip on brand-new, never-worked issues sitting at Actionable with no PR and no claim: #14558, #14559, #14554, #14557 (all opened in the previous ~2h by the scanner as repeat-failure split-outs). A little later #14525, #14550 and rationguard#396 (Claimed / PR in progress, updated 2m ago) picked it up too, while older rows at the same stages did not.
Owner: "actionable items that show 'regressed' are odd to me. makes me think I need to take action." and "shouldn't the regressed label come off the issue/pr if it is moving ahead?"
Root cause
The chip is driven by Record.RegressedFrom in the stage store (src/pkg/stagelabels/store.go), set by Store.Observe whenever IsRegression(prev, next) is true; stage_tracker.go turns rec.Regressed() into the regressed modifier. No GitHub label is involved (stage/regressed is retired).
IsRegression treated every numerically backward hop as a regression except a short exception list, so three ordinary hive moves were booked as falls:
The marker only cleared once the item climbed back to the exact stage it fell from, so it stuck through later forward moves.
The frontend always drew it red and counted it in Needs attention, regardless of whether anything waited on a person.
Fix
IsRegression: 020 is a reservation (leaving it is never a regression); 030 → 020 is not a fall; 040 → 030 is not a fall. A 040 → 010/015 fall is progress, not a regression, when the issue carries hive/partially-addressed (Decision.Progress / Store.ObserveProgress).
Self-clearing: the marker clears on the first forward hop of any size (also on a progress fall and when an old record's fall no longer counts under the current rule), in the same tick as the stage label write.
Dashboard: the regressed chip is red only on human-gated rows (human-gate modifier or stage 045/071/072/085/090). Otherwise it is a muted dashed stage ↓ pill with a tooltip naming the fall and saying the hive is handling it, yields the lead slot to other modifiers, and is excluded from Needs attention / the "Sent back" headline.
Lifecycle (set / kept / cleared) documented in labels-and-control-signals.md and dashboard-sections.md.
🔒 Claimed by @thatguywob (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.
PR #14571 (#14571) already covers all four requested changes: reservation/evidence-flap exceptions, partial-merge progress through both reconciliation paths, forward-hop and stale-record clearing, and human-gated dashboard presentation with lifecycle documentation and regression tests. Verified its current head e7ed0d6 and that it is open, ready for review, and targets v5. It is not merged or deployed yet. No additional implementation PR is needed for this dispatch.
@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
hivecommons-hive commentedon Oct 27, 2026
🔒 Claimed by @thatguywob (contributor on hive hosted-available-oke-11-placeholder-r05x). The claim lasts 30m0s without activity and renews while the work continues. Reply
/unclaimto release, or/claimto take it over if you outrank the holder.thatguywob commentedon Oct 27, 2026
PR #14571 (#14571) already covers all four requested changes: reservation/evidence-flap exceptions, partial-merge progress through both reconciliation paths, forward-hop and stale-record clearing, and human-gated dashboard presentation with lifecycle documentation and regression tests. Verified its current head e7ed0d6 and that it is open, ready for review, and targets v5. It is not merged or deployed yet. No additional implementation PR is needed for this dispatch.
— hive: backend=codex
🐝 Hive Agent:
contributor| SHA:unknowngithub-actions commentedon Oct 27, 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.