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
queen: a runner bee that committed nothing is an empty attempt after the grace, not a 6-hour wait (reviewer.t27 no_bundle_is_empty, behind TRIOS_QUEEN_NO_BUNDLE_EMPTY) #1886
Measured on production deploy 363860fd, 07:10-09:43Z on 2026-10-11:
425 of the reviewer's 480 row visits answered wait with "the diff of queen-N could not be read".
A runner stores its bee's branch in queen_bundle only when the bee committed something (queen-runner.tsstoreBundle). For a bee that committed nothing, two things follow:
importRunnerBranch answers ok with nothing imported;
committedFilesResult fails on a branch that does not exist.
The row is written wait, which is not counted against the issue, and is revisited at most every 15 minutes.
The issue stays claimed until WAIT_FROZEN_FLOOR_MS (6 h).
A container bee that commits nothing is written empty instead, and its issue is free after RETRY_FLOOR_MINUTES (30).
Requirements
FR-001: the review applies specs/queen/reviewer.t27no_bundle_is_empty at the point where it writes that wait row. The card, not the runtime, decides.
A finished runner row with no stored bundle, finished at least NO_BUNDLE_GRACE_MINUTES (10) ago, is judged exactly as a container bee that committed nothing.
Inside the grace it still waits.
FR-002: the behaviour is behind TRIOS_QUEEN_NO_BUNDLE_EMPTY=on, off by default. With the flag off, the review is unchanged, and a test proves it.
FR-003: the reviewer card is re-vendored byte-identical from t27, rebuilt with every fn exported, and PINned.
Success Criteria
Tests cover four cases: flag off, flag on inside the grace, flag on past the grace, and a bundle 40 s late that is judged on its branch.
Activity