Skip to content

πŸ› delegate: an offline author's PR is not "picked up" by someone else's activity - #18815

Merged
hivecommons-hive[bot] merged 4 commits into
hivecommons:v6from
Danathar:fix/delegate-rework-pickup-and-timed-out-wait
Nov 17, 2026
Merged

hivecommons-hive[bot] merged 4 commits into
hivecommons:v6from
Danathar:fix/delegate-rework-pickup-and-timed-out-wait

Conversation

@Danathar

Copy link
Copy Markdown
Contributor

Summary

"Let hive decide" sent a PR back to an author whose clanker had been offline for a day. It mistook a maintainer's review for the author picking the rework up. When nothing happened, the PR also dropped out of Waiting on you for good, so nobody was told.

This PR fixes both:

  • Only real evidence now counts as the author picking the rework up: their relay holding the fix task, or a new head.
  • A hive decision whose effect timed out shows in Waiting on you again.

What happened

Seen on projectbluefin/bluefin-bling#128, on hive build 3b50113. Times are UTC.

  • 11-08 20:44: the PR parks at 085 (sign-off attribution, the owner-designed park).
  • 11-11 15:03: a maintainer leaves a review comment. The author's last activity on the PR is 10-30, and their clanker was last seen 11-11 19:44.
  • 11-12 19:16: "Let hive decide" chooses return-to-contributor. Its comment on the PR says "the author's clanker has picked the rework up".
  • 11-12 19:46 onward: the row says "Sent back Β· no effect after 30m β€” options in the row menu". Nothing re-runs the decision, and the PR never appears in Waiting on you.
  • 11-17: a maintainer closes it by hand.

Cause

1. The pickup signal counted anyone's activity.

  • The return rung fires when ContributorOnline || ReworkPickedUp.
  • ContributorOnline was false: the author had no activity on the PR within the 15-minute presence window.
  • ReworkPickedUp comes from delegatedReworkPickedUp, and it counted the rework tracker's acknowledged state.
  • That state only means the PR's updated_at moved after the request (stage_tracker_rework.go:236-250, "somebody replied or edited"). The maintainer's review on 11-11 moved it.
  • The pickup signal was designed in Stage tracker: no signal whether the author or clanker has picked up a requested reworkΒ #15280 as activity by "the contributor or their clanker". updated_at cannot say who acted.

2. A timed-out decision left Waiting on you for good.

  • stageTrackerWaitKind skips every row that has a decision record ("a decided row is the hive's to act on next").
  • After the 30-minute effect timeout nothing re-runs the decision. Only the wording changes (stage_tracker_decided.go:206-208). So the row was nobody's to act on, and the badge, the pill and the MCP's waiting_on_you all left it out.

What changed

  • delegatedReworkPickedUp counts only:

    • in_progress: the author's relay holds the fix task;
    • pushed: a new head.

    With an offline author and neither of those, the ladder falls through to rung 5 (adopt in supersede mode), as Let hive decide adopts PRs whose issue was already fixed and merged (#15219 / #15240)Β #15289 designed.

  • stageTrackerWaitKind still skips a row with a pending or landed decision, but not one whose effect timed out. That row's wait kind comes from its real stage again (085 β†’ needs-decision, 090 β†’ ready-for-human-merge, and so on).

  • The Input.ReworkPickedUp comment and a changelog fragment are updated to match.

The tracker's own acknowledged chip on 071/072 rows is unchanged. It is a display hint ("something moved"), not a claim about who acted.

Not in this PR

Tests

  • delegated_rework_pickup_test.go replays Add in-progress issue pills and restart reset buttonΒ #128. Its first assertion checks that the tracker really reads the maintainer's review as acknowledged. It then checks:
    • the delegate does not count that acknowledged state as a pickup;
    • an in-flight fix task held by the author's relay does count;
    • a task that is only queued does not count.
  • stage_tracker_waiting_test.go: a timed-out decided row at 085 counts as needs-decision, and a decided row with no timeout still doesn't count.
  • stage_tracker_decided_test.go: the existing "timed out falls back to the real stage" case now also checks the row is a ready-for-human-merge wait.

Only gofmt -l was run locally (AGENTS.md bans local go test/go build/go vet); this PR's CI is the test run.

πŸ€– Generated with Claude Code

https://claude.ai/code/session_01BQYnPGFJx81c8JXR3o1sg1

…'s activity

"Let hive decide" returned projectbluefin/bluefin-bling#128 to an author
whose clanker had been offline for a day (rule return-to-contributor,
"the author's clanker has picked the rework up"). delegatedReworkPickedUp
counted the tracker's "acknowledged" rework state, which only says the
PR's updated_at moved after the request: a maintainer's review on 11-11
moved it. The direction timed out after 30 minutes, and because
stageTrackerWaitKind skips every row with a decision record, the PR
never reached Waiting on you; it sat five days until closed by hand.

- delegatedReworkPickedUp counts only in_progress (the author's relay
  holds the fix task) and pushed (a new head). An offline author now
  falls through to the adoption rung.
- stageTrackerWaitKind skips a decided row only while its effect has not
  timed out; a timed-out decision is a person's wait again by the row's
  real stage.

Signed-off-by: Danathar <6772335+Danathar@users.noreply.github.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BQYnPGFJx81c8JXR3o1sg1
@kubestellar-prow kubestellar-prow Bot added the dco-signoff: yes Indicates the PR's author has signed the DCO. label Nov 17, 2026
@kubestellar-prow

Copy link
Copy Markdown
Contributor

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by:
Once this PR has been reviewed and has the lgtm label, please assign hanthor for approval. For more information see the Code Review Process.

The full list of commands accepted by this bot can be found here.

Details Needs approval from an approver in each of these files:

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

@kubestellar-prow kubestellar-prow Bot added the size/M Denotes a PR that changes 30-99 lines, ignoring generated files. label Nov 17, 2026
@github-actions

Copy link
Copy Markdown
Contributor

Open PR overlap check

Base branch: v6

Conflicts are checked with every PR merged onto the current tip of v6, so a PR that only needs a rebase is not reported as conflicting with its siblings.

PR Title Relation Files
#18810 πŸ› automerge: an outdated change request waits on its reviewer, not the author same files src/pkg/dashboard/stage_tracker_waiting.go

@hivecommons-hive

Copy link
Copy Markdown
Contributor

@Danathar this PR is back in your court (stage/072-returned-to-contributor). The hive will not merge it until the head changes.

Failing checks:

  • check-cancelled

Reply here or push a new head to this branch (same sign-off); CI and the review swarm re-run automatically and the PR returns to the normal path.

Queued in your relay fix queue (2 waiting): fixes are handed out before new work as soon as you have capacity.

@hivecommons-hive

Copy link
Copy Markdown
Contributor

CI note: the only non-green check is mutation (Mutation (advisory), run 44162978012). It is an infrastructure failure, not a code failure: the actions/checkout step hung and was cancelled at the 20-minute mark, so setup-go, gremlins and the mutation report never ran. The trailing "Upsert mutation comment" failure (src/scripts/mutation-report.sh: No such file or directory) follows from the missing checkout. All other checks pass. No code change is needed; a re-run of the mutation workflow should clear it. Remaining gate is the tide lgtm/approved labels.


🐝 Hive Agent: scanner | Instance: hosted-available-oke-11-placeholder-r05x | SHA: unknown

β€” hive: agent=scanner backend=claude model=claude-sonnet-5-5 claude=2.1.326

Danathar and others added 2 commits November 17, 2026 20:36
The Mutation (advisory) job on 291d6d1 was cancelled at its 20-minute
timeout while actions/checkout's `git fetch` hung; no code ran. No
source change β€” this empty commit gives the PR a fresh head for CI.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Danathar <6772335+Danathar@users.noreply.github.com>
@hivecommons-hive
hivecommons-hive Bot merged commit 177db71 into hivecommons:v6 Nov 17, 2026
72 of 73 checks passed
@hivecommons-hive

hivecommons-hive Bot commented Nov 17, 2026 •

Copy link
Copy Markdown
Contributor

Hive status β€” updated in place; edits do not notify. Full history is in the Activity Monitor.

Evidence bundle `#188158eb2c5b` sha256:6526109b0c44510d0853eb6f40f1b9bb5c416055e097180596dc05a5df85f440 β€” download from the dashboard or `hivectl review evidence #18815`.

merged by Hive trusted-author auto-merge: author Danathar holds role owner and GitHub permission maintain on repo; CI green at 8eb2c5b.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

dco-signoff: yes Indicates the PR's author has signed the DCO. size/M Denotes a PR that changes 30-99 lines, ignoring generated files.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant