fix-from-logs

To fix bugs from CI failure logs, paste the raw error output to trigger a triage workflow that clusters related errors, localizes the root cause to file and line, and proposes a targeted fix with a typed pytest regression test.

Diagnosing a stack trace involves clustering related errors and pinpointing the root cause location, then entering plan mode to propose a minimal code fix and a failing pytest test before any changes are applied.

Yes, the workflow enforces a test-first discipline by writing a typed pytest regression test with mocks before implementing the code fix, ensuring the test fails initially and turns green after the patch.

This approach works specifically for bug triage from logs, stack traces, or CI failures, and does not trigger for general code review, refactoring, performance work, or test coverage requests without an associated failure.

Localizing a root cause from raw error logs involves analyzing the stack trace to pinpoint the exact file, function, and line number, then clustering related errors into actionable blocks for a targeted fix.

Plan mode is required before applying a patch from CI failure output to ensure the proposed fix and regression test are human-approved, minimizing risk by reviewing the root cause analysis and test strategy first.