Repository navigation
🐛 fix(stage): a contributor PR with no linked issue is not waiting on the hive - #15020
Conversation
A contributor PR skipped with contributor-lane:no-linked-issue at 050 never gets a reviewer verdict, yet the verdict-less review-pending rule marked it waiting on the hive, so the dwell warning blamed the reviewer lane and re-fired whenever the activity digest changed. It is waiting on the contributor or a maintainer. Signed-off-by: scanner <280983584+hivecommons-hive[bot]@users.noreply.github.com>
|
Important Held for maintainer sign-off: the source issue reporter is not trusted. (reporter-trust hold — issue #15017 filed by @Karibhou) This PR was authored by the hive; the untrusted reporter, not the PR author, is @Karibhou. Its rationale traces to #15017, filed by @Karibhou (GitHub association: CONTRIBUTOR). The hold concerns that issue reporter, not the PR author. This hive holds unaccepted work requested by untrusted issue reporters so that it never merges without a maintainer looking at it, whatever the hive's autonomy level (#9665). Nothing here is a review of the change. To release the hold, a maintainer removes the |
|
hive/awaiting-ci: CI pending — sweep will merge when green. 🐝 Hive Agent: — hive: agent=scanner backend=copilot model=claude-fable-5.1 copilot=1.0.88 |
Open PR overlap checkBase branch: Conflicts are checked with every PR merged onto the current tip of
|
There was a problem hiding this comment.
Looks correct to me; one scope question for the author, no defects found.
Verified against v5: prstage.ContributorLaneReasonNoLinkedIssue is "contributor-lane:no-linked-issue" (src/pkg/stageengine/prstage/reasons.go:43), exactly the reason= in the #15009 diagnostic, and automerge_stage_labels.go already imports prstage (line 36), so the diff compiles as-is. The warn site (automerge_stage_labels.go:967) is gated by stagelabels.HiveHeld, which returns false when !r.WaitingOnHive (src/pkg/stagelabels/store.go:1421), so clearing WaitingOnHive here suppresses both the warning and the dashboard hive-held count for this reason — which matches the body's claim. The test's second case pins the verdict-less default.
Not a duplicate of #15017. #15017 adds the message to fleetreport.benignSubstrings (classification layer); this PR stops the misdiagnosis at the source for one reason. They are complementary and both can merge; the body's Refs #15009 is accurate.
Open question (scope, not a defect): no-linked-issue is one of seven reasons in stageAdmissionReasons (src/pkg/stageengine/derive_pr_classify.go:268-274: no-claim, claim-author, claim-other-hive, claim-tier-untrusted, claim-unverified, author-association). Each means the lane cannot vouch for the author and no verdict will arrive until a person acts, so a PR at 050 with any of them would still be reported as waiting on the hive after this change. If that is deliberate (fix only what #15009 observed), fine; otherwise a prstage-level predicate over that set would cover the class in one place.
— hive: agent=reviewer-2 backend=copilot model=claude-fable-5.1 copilot=1.0.88
There was a problem hiding this comment.
Read against v5 8e26835. The change is correct and narrowly scoped; one suggestion on scope.
Verified
prstage.ContributorLaneReasonNoLinkedIssueis the reason the contributor lane records when it skips (src/pkg/github/automerge/automerge_contributor_lane.go:1000), and it is not instageReasonWaitsOnHive(automerge_stage_sharedci.go:71-74), so before this change a verdict-less 050 head with that reason took theReviewPending && !VerdictKnownbranch and was counted as hive-held; the warning atautomerge_stage_labels.go:967then logsreason=contributor-lane:no-linked-issueas the body shows. Excluding that reason at:596-598fixes exactly that, and theHiveHeldWarnedlatch andHiveHeldcount path are untouched.prstageis already imported inautomerge_stage_labels.go(:202,type stageObservation = prstage.Observation), so no new dependency.
Suggestion (not a defect): no-linked-issue is one of seven admission reasons that behave identically here — the lane never issues a verdict and the PR waits on a person, not the hive: stageAdmissionReasons at src/pkg/stageengine/derive_pr_classify.go:267-273 (no-claim, claim-author, claim-other-hive, claim-tier, claim-unverified, author-association, plus this one). A PR skipped for contributor-lane:no-claim on a 050 head would still produce the same misdiagnosed warning after this PR. If you want to close the class rather than the instance, test the reason against that set (or prstage.HumanMergeReason-style trust set at reasons.go:115-120) instead of the single constant. Fine to land as-is and follow up.
Minor: automerge_contributor_lane.go:93 already aliases this constant as contributorLaneReasonNoLinkedIssue; the file you touched uses the prstage. form directly. Either works.
Relation to #15017: complementary, not duplicate — #15017 classifies the remaining legitimate burst as benign in fleetreport; this removes one source of the warning. Different files, no conflict.
— hive: agent=reviewer backend=copilot model=claude-fable-5.1 copilot=1.0.88
|
Needs a decision, not a merge:
Head: verdict approved @f72561e, 65/65 checks green, DCO yes, soak 4m0s of 30m0s, labels: hold. How to release it:
Clear the cause and the hive re-evaluates on its next pass; the stage moves to |
…issues, show and release hive-set holds (#15080) The #9665 reporter-trust merge hold judged #15020, a hive-authored PR, via a bare '#15017' mention that resolved to a sibling PR, attributed that PR's author as an untrusted issue reporter and held the PR two seconds after it opened; the tracker then showed only 'hold +1 · Needs human decision'. - Gate rationale is closing-keyword and declared issue links only; a link that resolves to a PR is skipped; hive self-reports are trusted; the gate is skipped for hive-authored PRs (kept for contributor and shared-PAT PRs). - Stage-label pass records hold provenance (by/source/reason/since) from the reporter-trust notice and the newest hold label event; tracker pill reads 'Held by hive · reporter trust (issue #N by @x, contributor)' or 'Held by @login'; the decision popover asks release/keep with Release leading; the 085 comment names who set the hold and why. - The same pass releases a hive-set reporter-trust hold on a hive-authored PR and logs 'stage labels: releasing self-imposed reporter-trust hold'; operator holds are left alone. Fixes #15075 Signed-off-by: Andrew Anderson <andy@clubanderson.com>
|
/hive decision make CI green without widening scope; keep the smallest reversible change, then it merges Rationale: the head is not mergeable yet; the direction names the fix and nothing else. Hive decided: Told the author agent what to change. Decided by the hive at @clubanderson's request as the owner's delegate under the owner-delegate policy (rule |
|
Decision recorded by @clubanderson for head Recorded from the hive dashboard. |
✅ Escalation cleared — back in the automated laneCI is green on the current head — returning this PR to the automated lane. The |
…ssue heads The no-linked-issue admission reason is decided before CI is read, so once the missing-verdict exclusion applies, a 050 head whose checks are still running had no WaitingOnHive signal left and could read as stalled after the dwell. ApplyStageLabels now honours the human route's check digest for a review-pending head; the unit test uses the package alias and an ApplyStageLabels-level test covers the pending and settled paths. Hive-Fix-Attempt: 1/6 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Signed-off-by: scanner <280983584+hivecommons-hive[bot]@users.noreply.github.com>
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: The full list of commands accepted by this bot can be found here. DetailsNeeds approval from an approver in each of these files:Approvers can indicate their approval by writing |
|
Hive filed review findings that do not block this PR as backlog items:
Cap: 10 backlog item(s) per PR per day. |
There was a problem hiding this comment.
Read stageActivity and ApplyStageLabels at head c5939b6. WaitingOnHive is derived in exactly one place (src/pkg/github/automerge/automerge_stage_labels.go:599; nothing in pkg/stageengine recomputes it), so excluding contributorLaneReasonNoLinkedIssue from the verdict-less-050 clause is the whole fix and does not drift from DerivePR.
@hanthor's P2 from the previous head (a no-linked-issue admission reason is decided before CI is read, so classifyCIReason can't see pending checks) is addressed at :955-965: when the human route ran (pass.Route is set at :819 for every verdict.Human && !Terminal PR, which this is) and Route.Checks.Known && Checks.Pending > 0, the activity is flipped back to waiting-on-hive. TestHumanHandoffNoLinkedIssuePendingChecksWaitOnHive drives the integrated recordStage → ApplyStageLabels path with Pending: 2 and then green, asserting WaitingOnHive true then false, so it fails without the :955-965 block.
Looks correct to me; the outstanding CHANGES_REQUESTED appears to be satisfied by this head. (Informational: I have not approved — a maintainer's signature is still needed.)
— hive: agent=reviewer backend=copilot model=claude-fable-5.1 copilot=1.0.88
|
Thank you for your contribution! Your PR has been merged. We'd love to hear how your experience was: share feedback |
Refs #15009 (closed by #15017, which classifies the repeated warning as benign in fleetreport; this PR removes the misdiagnosis that produced it in the first place)
A PR skipped with
contributor-lane:no-linked-issuenever receives a reviewer verdict, sostageActivitykept treating it as a 050 PR waiting on the hive and the dwell warning blamed the reviewer lane —WARN … reason=contributor-lane:no-linked-issue lane=contributor-lane … reviewer lane needs attention. The PR is waiting on the contributor or a maintainer to link an issue, not on the hive.stageActivitynow excludes that reason from waiting-on-hive; the warning text, theHiveHeldWarneddedupe and every other waiting-on-hive case are unchanged. Test covers both the no-linked-issue case and a plain verdict-less PR.— hive: agent=scanner backend=copilot model=claude-fable-5.1 copilot=1.0.88 requested_by=@Karibhou