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
Decide Whether #2046 Merges Over Unstated Copilot Coverage #2048
Handoff: #2042, track auto-1904, for #1904. This decision blocks pull request #2046 (feature/auto-1904 -> develop, head c0cf4c2b).
Question
Should #2046 be merged to develop as it stands, even though Copilot's review does not state its coverage?
The state of the pull request:
Two Copilot rounds cover head c0cf4c2b, the first one self-seeded and the second one re-requested. Both are ccr-overview-v2 overviews with Findings: None.
There are no threads, and suppressed shows on_head=0. All 10 of 10 checks are green, and the merge state is CLEAN.
pr_review.py wait and status exit 45 COVERAGE_IS_UNSTATED (overview=0/0). The overview has no "reviewed N of M files" statement, so Merge Gate item 2 is not satisfied, and merging over it is the maintainer's decision.
Round 1's "What changed" file table names all 8 changed files. With its zero-width spaces stripped, comm -3 against gh pr view 2046 --json files shows no difference.
Both rounds put the verdict under "Needs a closer look". The reason they give is the PR-review tooling and the carried governance text it changes, and the GitHub API timing that cannot be verified from the diff. They do not name a defect.
The local strict review pass on this head found nothing, and it is recorded.
The root cause of this shape is open in #2047. It is the same shape as the per-PR decisions #2034, #2037 and #2040. None of those answers carries over to this pull request.
Options
Recommended: squash-merge #2046 as it stands. The file table shows every changed file was read, two rounds found nothing, and the local pass found nothing. The fix for #1904 lands on develop now, and the lane is closed out after the merge.
Request a third Copilot round and wait for one that states coverage. This delays the merge by at least one round. Two re-requested rounds have already come back in the same shape, and #2047 records that re-requesting does not clear it.
Review #2046 yourself before it merges. This answers the "Needs a closer look" verdict, which is aimed at the timing behavior. The branch and pull request stay open until you merge or comment.
Activity
ptr727 commentedon Sep 29, 2026
Answered by the maintainer on 2026-09-29: option 1, squash-merge #2046 as it stands.