qa-loop

name	qa-loop
description	Automated test-fix-retest cycle. Runs tests, fixes failures, and re-runs until all pass or a maximum retry count is reached. Activate when #qa appears anywhere in the user message.
argument-hint	[test-command-or-scope]
disable-model-invocation	false
allowed-tools	Read, Edit, MultiEdit, Write, Glob, Grep, Bash, Task

Run QA loop for: $ARGUMENTS

Execute an automated test → fix → retest cycle.

Step 1 — Identify the test command:

• If the user specified a command, use it.
• Otherwise, ask test-commander for the narrowest useful verification.
• If no test infrastructure exists, report what manual checks are needed and stop.

Step 2 — Run the initial test:

• Execute the test command.
• Capture full output including error messages, stack traces, and exit code.

Step 3 — Analyze failures:

• For each failure, identify:
  • Which test failed and what it expected.
  • The likely root cause (use bug-hunter if unclear).
  • The minimal fix needed.

Step 4 — Fix and retest:

• Apply the smallest fix for each failure.
• Re-run the test command.
• If new failures appear, analyze and fix those too.
• Maximum 5 fix-retest cycles to prevent infinite loops.

Step 5 — Report results:

• Track: iteration number, tests run, passed, failed, fixed.
• If all tests pass, report success.
• If failures remain after max cycles, report what is still broken and why.

Escalation:

• After 2 consecutive failures on the same test, bring in bug-hunter for deeper analysis.
• After 3 cycles with no progress, bring in oracle for architectural guidance.
• If a fix would require scope expansion beyond the original change, stop and report.

End with:

• Test command
• Iterations (number of cycles run)
• Result (all pass, partial, or blocked)
• Fixes applied (file and description for each)
• Remaining failures (if any, with analysis)