Test Runner
Rule: A change that breaks existing tests is not done. Run tests after every code change, not at the end.
Steps
Run tests:
bash ~/.claude/skills/test-runner/scripts/run-tests.shFor coverage report:
bash ~/.claude/skills/test-runner/scripts/run-tests.sh --covFrom the output, identify:
- Failing tests — fix before continuing, never skip
- New code with no tests — flag it explicitly
- Coverage drop — if it decreased, justify it or add tests
After Fixing a Failure
Re-run immediately after each fix — never batch multiple fixes and run once at the end.