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
Part of epic #1184 (PM stack coherence review, 2026-10-26).
Split from #1162's scope. The in-flight lane dedupe is deliberately coarse: one shared ref or a title match drops the whole orchestrator entry, including its other refs (live.ts inFlightBoardItems). Real multi-ref work can therefore vanish from the board when a Kaneo card shares one ref.
Acceptance:
Dedupe is per-ref instead of per-task.
A multi-ref task deduped on one ref still shows with its remaining refs.
Render tests cover the multi-ref dedupe case.
Acceptance criteria
Board dedupe of orchestrator tasks is per-ref instead of per-task.
A multi-ref task deduped on one ref still shows on the board with its remaining refs.
Notes / risks: Per-ref dedupe changes what the board shows for every multi-ref orchestrator entry — render tests should include the single-ref case to prove no regression.
Suggested approach: Work in live.ts inFlightBoardItems; keep the coarse title-match behaviour out — only ref-level dedupe.
Live check: verified still open and unheld this run; a merged-PR search on this run (2026-10-28T10:35:00Z) returned no PR resolving this issue, so it remains live work. Acceptance criteria appended to the body (original text preserved verbatim above it) and labels normalized by the fleet monitor.
Triage verdict: confirmed, route to implementation
At main f00dde6, inFlightBoardItems discards a whole task when any ref matches an existing card, hiding unmatched refs. The existing board-in-flight test asserts this coarse behavior. Filter matching refs instead, preserve genuinely ref-less tasks, remove title-only cross-card dedupe, and retain the remaining refs and their proper repository. Add regression tests including rendered links, fully matched single refs, ref-less tasks, distinct-ref title matches and section repeats. No relevant linked/open PR was found in issue-number, function-name, or per-ref searches.
Queue task 4901 is complete with PR #3934 open: #3934. The focused board-in-flight suite passes; broader unrelated baseline failures are recorded in the task result.
Activity
developing-today commentedon Oct 27, 2026
CODE-330
dezren39 commentedon Oct 28, 2026
Intake review
Related: Part of epic #1184; split from #1162. Sibling batch: #1198, #1197, #1195.
Notes / risks: Per-ref dedupe changes what the board shows for every multi-ref orchestrator entry — render tests should include the single-ref case to prove no regression.
Suggested approach: Work in live.ts inFlightBoardItems; keep the coarse title-match behaviour out — only ref-level dedupe.
Live check: verified still open and unheld this run; a merged-PR search on this run (2026-10-28T10:35:00Z) returned no PR resolving this issue, so it remains live work. Acceptance criteria appended to the body (original text preserved verbatim above it) and labels normalized by the fleet monitor.
1 remaining item
dezren39 commentedon Nov 1, 2026
Triage verdict: confirmed, route to implementation
At main f00dde6, inFlightBoardItems discards a whole task when any ref matches an existing card, hiding unmatched refs. The existing board-in-flight test asserts this coarse behavior. Filter matching refs instead, preserve genuinely ref-less tasks, remove title-only cross-card dedupe, and retain the remaining refs and their proper repository. Add regression tests including rendered links, fully matched single refs, ref-less tasks, distinct-ref title matches and section repeats. No relevant linked/open PR was found in issue-number, function-name, or per-ref searches.
dezren39 commentedon Nov 1, 2026
Triage complete: confirmed on current main and queued implementation follow-up task 4901. Issue remains open for the fix.
dezren39 commentedon Nov 1, 2026
Picked up queue task 4901 for the board per-ref dedupe fix. Fresh check: issue #1196 remains open; relevant open-PR searches found no competing fix.
dezren39 commentedon Nov 1, 2026
Queue task 4901 is complete with PR #3934 open: #3934. The focused board-in-flight suite passes; broader unrelated baseline failures are recorded in the task result.
Merge pull request #3934 from developing-today/queue/4901-1196-per-re…