# slop-mop

> A code-quality gate system for AI coding agents. It catches the shortcuts agents take under reward pressure — tests that assert nothing, dead code, silenced quality gates, and `--no-verify` commit-hook bypasses — then hands back concrete cleanup work instead of a scolding.

slop-mop runs as a CLI (`sm`) that any coding agent can call, ships as a Claude plugin and skill, and runs in CI via GitHub Actions. It is free and open source. Install with `pipx install slopmop[all]`, then run `sm swab`.

## Commands
- [swab](https://github.com/ScienceIsNeato/slop-mop): the fast pass — run after every change; catches the obvious slop (tests that assert nothing, oversized functions, secrets in fixtures)
- [scour](https://github.com/ScienceIsNeato/slop-mop): the deep pass — every gate at once before you open or update a PR, including a dependency/CVE audit
- [buff](https://github.com/ScienceIsNeato/slop-mop): work the review — turns CI and review feedback into the agent's sequenced next step
- [sail](https://github.com/ScienceIsNeato/slop-mop): runs the right verb for you when you're not sure what comes next
- [refit](https://github.com/ScienceIsNeato/slop-mop): a structured cleanup pass for a repo you've inherited
- [barnacle](https://github.com/ScienceIsNeato/slop-mop): file friction upstream when the tool is wrong, instead of routing around it
- [gang press](https://github.com/ScienceIsNeato/slop-mop): one-time manual setup a human runs — installs shell intercepts and a git wrapper so reflex commands (pytest, gh, mypy) route to sm and `git commit --no-verify` is blocked

## Install
- [PyPI — slopmop](https://pypi.org/project/slopmop/): `pipx install slopmop[all]`
- [Use with Claude](https://github.com/ScienceIsNeato/slop-mop#use-with-claude): plugin and skill, auto-triggers on remediation, six slash commands
- [slop-mop-action](https://github.com/ScienceIsNeato/slop-mop-action): run the gates in GitHub Actions

## Docs and source
- [GitHub — ScienceIsNeato/slop-mop](https://github.com/ScienceIsNeato/slop-mop): the tool, gates, and docs
- [FAQ](https://www.slop-mop.com/faq): common questions about slop-mop

## Evidence
- [Case studies](https://www.slop-mop.com/case-studies): real onboarding walkthroughs, with what each run fixed versus baselined
- [rulebook-ai, cold](https://www.slop-mop.com/case-studies/rulebook-ai): ~8.5k-line repo taken from 8 failing gates to A+; found two latent crashes and a CI pipeline dead since September 2025
- [Onboarding OpenHands](https://www.slop-mop.com/case-studies/openhands): ~322k lines, mostly baselining — the honest ceiling for a codebase that size
- [In the wild](https://www.slop-mop.com/in-the-wild): real captured agent sessions
- [Harm Reduction for Addicted Agents](https://scienceisneato.substack.com/p/slop-mop-harm-reduction-for-addicted): why it exists and why it works