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
[deep-report] Add repeat-offender detection to the automated [aw] workflow-failure issue filer #48672
Across this week's data, the same workflow failing repeatedly gets a brand-new siloed [aw] <workflow> failed/produced no safe outputs issue every time instead of being linked/consolidated: Matt Pocock Skills Reviewer (9+ issues since 07-25), Design Decision Gate (3+ issues), and Daily Container Image Security Scan (4 issues: #47939, #48152, #48311, #48549) all show this pattern. This mirrors the same anti-pattern DeepReport's own prior run (#47811) had to manually clean up for goroutine/timeout fixes.
Expected Impact
When the automated failure detector is about to file an issue for a workflow that already has an open failure issue from the last N days, it should comment/re-open the existing issue instead of creating a new one. This directly reduces backlog noise and makes chronic failures visible as a single trend line instead of N disconnected data points.
Suggested Agent
New Agent (or extend the existing [aw] failure-detection automation).
Activity
github-actions commentedon Jul 28, 2026
🍪 Issue Monster selected this for Copilot
I have identified this issue as a good candidate for automated resolution and requested assignment to the Copilot coding agent.
If assignment succeeds, the Copilot coding agent will analyze the issue and create a pull request with the fix.
Om nom nom! 🍪
github-actions commentedon Jul 30, 2026
This issue was automatically closed because it expired on 2026-07-30T15:41:43.027Z.