How does this skill identify what to remove?

It performs a git diff between your current branch and the main branch to isolate recent changes, then evaluates those changes against common AI-generated patterns and your file's existing style.

Will it accidentally delete my legitimate logic changes?

The skill is specifically programmed to target style inconsistencies and redundant patterns. It reports a summary of changes, allowing you to verify that your functional code remains intact.

Does it support languages other than Python?

Yes. While it has specific logic for Python (like moving inline imports), the principles of removing comments and defensive 'slop' apply to all major programming languages.

What is considered 'AI slop' in a codebase?

AI slop refers to unnecessary artifacts added by AI models, such as redundant comments that explain the obvious, overly defensive code in safe paths, or lazy type casts like 'any' used to bypass compiler checks.

AI Code Deslopper

Name: AI Code Deslopper
Author: davila7

bydavila7

•

16,312

•

Developer Tools

Refines AI-generated code by removing redundant comments, unnecessary defensive checks, and style inconsistencies to ensure human-grade quality.

The deslop skill for Claude Code streamlines the transition from AI-assisted drafting to production-ready code. It identifies and removes 'AI slop'—the characteristic artifacts left behind by LLMs, such as overly verbose comments, excessive try-catch blocks in trusted paths, and lazy type casting. By comparing your current branch against the main branch, it automatically standardizes your codebase, ensuring that AI-generated contributions blend seamlessly with your team's existing style and quality standards before you submit a pull request.

Key Features

01Refactoring of inline Python imports to standard top-level placement

02Standardization of type casts and removal of shortcut 'any' types

0316,312 GitHub stars

04Removal of redundant or non-human-like AI comments

05Elimination of unnecessary defensive checks and try-catch blocks

06Automated diff analysis against the main branch

Use Cases

01Cleaning up rapid prototypes generated by Claude or other AI agents

02Preparing an AI-assisted feature branch for a professional peer review

03Standardizing code style after a bulk refactor using AI tools

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add davila7/claude-code-templates sentry

For use in Claude.ai and ChatGPT

Download Skill