| name | mutmut-report |
| description | Use when the user wants to triage a Python project's surviving mutmut mutants after a run — classifies each survivor as REAL_GAP, EQUIVALENT, or UNTESTABLE and reports the test fixes worth making. Triggers — "which mutants survived", "triage mutation testing results". Do NOT use to run mutmut itself, for coverage reports (see coverage-gaps), or for non-Python mutation testing. |
| argument-hint | --module MODULE --max N |
| user-invocable | true |
mutmut Survivors Report
Run the analysis script against a project's existing mutants/ workspace, then report which surviving mutants are real test gaps and which are safe to accept.
CANARY: mutmut-report-loaded-2026-06-02-87a56d2e-e3c5c50abfce45f3
Step 1 — Run the script
Invoke from the project directory that owns the mutants/ workspace (the one where mutmut run was executed). $ARGUMENTS is forwarded verbatim to the script — pass flags like --module validate, --max 20, or --venv .venv.
python3 "${CLAUDE_SKILL_DIR}/scripts/analyze_mutmut.py" $ARGUMENTS
If the target project is a different checkout, cd to it first — the script reads mutants/ from the current directory while still being launched by its absolute ${CLAUDE_SKILL_DIR} path.
Step 2 — Present results
After running, report concisely:
- Which modules have Real gaps and how many
- For each real gap: what the fix is (tighter assertion, missing test case, etc.)
- Which are Equivalent or Untestable — just confirm these are accepted, no detail needed
- Suggest the top 2-3 fixes to implement next
Do not re-explain what mutmut is. Just report findings and fixes.
