| name | test-coverage-loop |
| description | Adds tests until the suite reaches 100% coverage. Drafts candidate tests in a throwaway worktree and reports them โ never commits a test file itself. |
test-coverage-loop procedure
The prompt carries intent; this skill carries the procedure โ keep it a checklist,
not an essay. Adapted from Forward Future Loop Library #5 (Matthew Berman) to
this library's L1-first, report-only shape.
- Read โ the coverage report. Identify uncovered regions not yet in
test-coverage-loop-state.md's register.
- Do one unit โ for up to 5 uncovered regions this beat (largest
first): in a fresh throwaway worktree, draft a test with a real
assertion (not just line execution), run the full suite.
- Verify:
- Coverage rose, suite green โ write a
coverage-report.md entry:
region, the drafted test, the coverage delta, the suite result.
- Still uncoverable or suite red after one retry in a clean worktree
โ write an
exclusion-candidate entry instead (region, why it resists
testing) and do not retry it again this cycle.
- Record โ update
test-coverage-loop-state.md's per-region status,
append one line to loop-run-log.md, commit both. Discard the throwaway
worktree.
- Take NO other action โ never commit a test file. This loop is L1
report-only until a human promotes it (see
LOOP.md โ the source's own
caveat on assertion quality applies at every level).