Will it break my project's architecture?

No. The Refactor skill is designed for tactical improvements within your existing architecture. For major structural changes, you should use the architectural review skill instead.

How does it ensure the refactored code still works?

Every change is passed to a QA agent that runs your test suite, linters, and any custom verification steps you provide. If a change fails, it is automatically reverted.

What does the Refactor skill actually do?

It runs an autonomous loop that scans your code for quality issues like complexity and duplication, implements fixes using AI specialists, verifies them with tests, and commits the results.

Can I control how aggressive the refactoring is?

Yes, you can choose between Maximum, High, and Low aggression levels to determine if the tool should attempt complex logic consolidations or stick to safe formatting and dead code removal.

Does it update my documentation?

Yes, after completing the refactoring loop, it automatically triggers a documentation review workflow to ensure your READMEs and docs stay in sync with the updated code.

Refactor - Autonomous Code Quality Improvement

Name: Refactor - Autonomous Code Quality Improvement
Author: chrisallenlane

bychrisallenlane

•

开发者工具

Automates the iterative process of scanning, refactoring, and verifying code quality improvements to reduce technical debt and enhance maintainability.

The Refactor skill provides an autonomous workflow for systematically cleaning up codebases through iterative improvements. It identifies tactical opportunities such as removing dead code, consolidating duplicate logic (DRY), and simplifying complex functions while strictly adhering to your existing architecture. Each refactoring batch is implemented by specialized agents, verified against your test suite by a dedicated QA agent, and committed atomically to Git. This ensures that your code becomes cleaner and more comprehensible without introducing regressions or altering module boundaries, making it an essential tool for continuous codebase health and technical debt management.

主要功能

01Autonomous iterative scanning and refactoring loops

02Language-specific SME agents for precise implementation

03Integrated QA verification with automated test and lint execution

045 GitHub stars

05Configurable aggression levels from 'Safest' to 'Aggressive'

06Atomic Git commits with detailed, descriptive messages

使用场景

01Automating routine code cleanup and maintenance before major feature releases

02Systematically removing dead code and duplicated logic across large projects

03Reducing technical debt in legacy codebases without changing core architecture

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add chrisallenlane/claude-plugins refactor

For use in Claude.ai and ChatGPT

Download Skill