Coverage.py
3,000Code coverage measurement for Python. Detects test-manipulation by surfacing uncovered code.
Maturity tier details
Computed tier: 2A(raw score 4.89). Effective tier: 2A.
Facets:documented_in=0evaluated_on=0integration_surfaces=1.5paper_citations=0recent_commit=2stars=1.39
Source of truth:pipeline/maturity/score.py
Description
Coverage.py instruments Python execution to record which lines and branches are exercised by a test run, then reports per-file and per-line coverage. By surfacing uncovered code it makes test_manipulation (deleted tests, skipped assertions, "tests pass but exercise nothing") visible, and exposes incomplete_execution where new code ships without any tests at all. Configuration lives in `pyproject.toml [tool.coverage.*]`, which the audit detector now recognises.
Why it's included
Occupies detection × post_generation × fully_autonomous for test_manipulation in the Python lane — distinct from pytest because the signal is "what was *not* run," not "what failed." Including it lets the audit detector resolve `[tool.coverage.*]` blocks and credit projects with a coverage-gating discipline.
Security notes
Standard dev-tool security posture; Coverage.py uses sys.settrace / sys.monitoring to instrument the Python interpreter, so it should not be enabled in production paths. Output reports may include source snippets — treat coverage artefacts with the same access controls as the source repo.
Addresses failure modes
- Test manipulation — Agent gaming its own evaluation — editing tests or test harness so failures disappear without fixing the underlying bug.
- Incomplete execution — Agent claims completion while subtasks are skipped, stubbed, or silently dropped.