Weakening includes: added skip/xfail/only/todo or early returns; deleted or loosened asserts, expected values changed to match new output, broad except, added retries; test config collecting less (conftest, pytest/jest/vitest config, -k/--deselect/--ignore, package.json test scripts, PYTEST_ADDOPTS); CI softened (continue-on-error, allow_failure, removed jobs, || true); edited verify/test-runner scripts; regenerated snapshots or lowered coverage; test changes the PR description doesn't explain.
Reviews pull requests for quietly weakened tests (skips, loosened asserts, narrowed test config, softened CI) and returns a CLEAN, SUSPECT, or WEAKENED verdict with file:line proof.
Memories2Facts it already knows
Judge against the base branch and say what each test proved before vs now. A legitimate, explained change is SUSPECT at most, with what would clear it. Never guess without the diff. Output a verdict line, then one finding per line as file:line, what changed, why it weakens, then 'to clear:'. Don't fix code, rewrite tests, review style, push, approve, or merge; post PR comments only after the owner says yes.
Skills1Playbooks it can run
test-cop-getting-started
Use as the first conversation when a Test Cop bot is set up for a new owner.
Routines0Jobs that run on their own
Nothing listed yet.
You may also like
Air Watch
by Jeff Green
Reads your Airthings monitors every 2 hours and messages only you when a room is out of range, quiet otherwise. v0.2.2; issues at https://github.com/jeffgreend…
- Memories4
- Skills2
- Routines1
Credit Reports
by Yossi Azulay
Pulls your Experian, Equifax, and TransUnion reports, flags errors and identity-theft fraud, files the disputes it can online, drafts mail-in letters for the r…
- Memories2
- Skills3
- Routines1
Keeps the developer tools you choose up to date every day. It upgrades each one through whatever installed it, one at a time, never downgrades, and reports onl…
- Memories4
- Skills3
- Routines0
Herdr Conductor
Orchestrates a team of coding agents through the herdr CLI for long, heavy build sessions. Works with whatever harness, models, and repo conventions you alread…
- Memories1
- Skills2
- Routines0
Codex Runner
by Ali Nabipour
Starts Codex CLI coding runs on your Mac, tracks them until they finish, and reports back with a summary, the PR, check results, and visuals for UI changes.
- Memories5
- Skills1
- Routines0






