grp06

grp06 / slop-janitor

Public

Repeatable multi-turn Codex refactor loop. ⭐️ star if you like it! ⭐️

45
1
69% credibility
Found Mar 23, 2026 at 45 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

slop-janitor is a tool that automates loops of AI planning, implementation, and review to refactor and enhance software code repositories.

How It Works

1
🔍 Discover slop-janitor

You find a handy tool that uses AI to automatically clean up and improve your messy code projects.

2
📥 Get everything ready

Download the cleanup tool and the AI helper it works with, then tell it where to find the AI part.

3
🔐 Sign into your AI account

Log in with your AI service account so the tool can use smart thinking on your behalf.

4
🧹 Start the cleanup in your project

Go to your project's folder, pick a mode like 'refactor' or give a goal like 'make it simpler', and launch the tool.

5
⏳ Watch it iterate and improve

The tool runs repeated loops of planning changes, refining ideas, making updates, and reviewing results to make your code better.

6
📋 Check the full story

Look at the detailed log file that records every step, token use, and what got changed.

✅ Celebrate cleaner code

Your project is now simpler, more reliable, and easier to work with, sometimes even auto-saved as checkpoints.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 45 to 45 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is slop-janitor?

Slop-janitor is a Python CLI that automates repeatable multi-turn Codex loops to clean slop from your codebases—like a janitor slop sink sucking out complexity from the janitor closet. Run it from any repo with `slop-janitor --mode refactor` to find high-leverage refactors, plan them via Codex exec plans, iterate improvements, implement changes, and review results on a single thread for continuity. It logs full runs, optionally auto-commits git changes, and requires a Codex workspace clone plus login.

Why is it gaining traction?

It skips manual prompt chains by bundling Codex skills for planning, improving plans four times by default, implementing, and reviewing five times per cycle—making refactors reliable without resetting context. Configurable cycles, prompts like "focus on testability," and detailed run logs in `runs/` let you inspect token usage and failures post-run. Developers star it for turning slop into solid code via janitor sink slop hopper type workflows.

Who should use this?

Backend Python devs refactoring legacy services bogged down by janitorial slop. Indie hackers prototyping CRMs or tools with Codex, needing multi-turn loops without babysitting chats. Teams evaluating AI for code cleanup before upstream merges.

Verdict

Try it if you have Codex access—solid docs, tests, and Python packaging make setup straightforward despite 45 stars and 0.7% credibility score signaling early maturity. Star if you like repeatable refactor loops, but pair with manual reviews until more battle-tested.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.