parallel-ci-triage

You can fix failing GitHub Actions jobs in parallel by splitting them into independent workstreams. Each failing job gets its own subagent to inspect logs, apply code changes, and verify the fix locally before merging.

Parallel triage for CI repair groups failed jobs into independent error clusters so unrelated issues do not block each other. Each cluster is handled by an isolated subagent that reviews logs, diagnoses root causes, and validates fixes locally.

GitHub CLI access is required to debug multi-job CI runs with parallel subagents, though copied logs also work. You must also be able to rerun the same failing commands locally to verify each fix before merging.

Yes, parallel subagents work for lint, test, and build pipeline failures. The triage process isolates each failing job so independent errors across different pipeline stages can be diagnosed and resolved simultaneously.

Avoid parallel debugging for failed CI jobs when failures are dependent rather than isolated. If one job's error triggers subsequent job failures, a sequential debugging pass is more appropriate than splitting into parallel workstreams.