What types of code smells can this skill identify?

It identifies long functions, excessive parameters, duplicated logic, deep nesting, and poorly named variables or classes.

Can I add new features while refactoring with this skill?

No, this skill enforces a strict separation between refactoring and feature development to ensure each change is isolated and safe.

Does this skill change how my code works?

No, the primary rule of this skill is to improve the internal structure of the code without changing its observable external behavior.

How does it handle large-scale structural changes?

It breaks down large refactors into small, incremental, and verifiable steps, running tests after every change to prevent regressions.

What happens if my code doesn't have existing tests?

The skill will guide you to write characterization tests first to capture current behavior as a safety net before any refactoring begins.

Safe Code Refactoring

Name: Safe Code Refactoring
Author: 686f6c61

by686f6c61

•

开发者工具

Refactors source code safely by identifying code smells and applying structural improvements while maintaining behavioral consistency through automated testing.

This skill provides a rigorous framework for refactoring code without altering its external behavior. It guides Claude through identifying common code smells—such as long functions, excessive parameters, and deep nesting—and applies transformations like function extraction and expression simplification. By enforcing a 'hard-gate' policy where tests must pass before and after every incremental change, it ensures that code improvements remain safe, maintainable, and free of regression bugs, effectively separating structural cleanup from feature development.

主要功能

0137 GitHub stars

02Strict separation of refactoring and feature logic in commits

03Automated code smell identification including duplication and complexity

04Test-driven safety gates to ensure zero behavioral regressions

05Incremental refactoring steps to ensure changes are reversible and verifiable

06Characterization test generation for legacy code without existing coverage

使用场景

01Simplifying complex conditional logic or deeply nested loops into readable patterns

02Reducing technical debt in legacy modules without breaking existing functionality

03Improving code maintainability and readability before starting new feature development

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add 686f6c61/alfred-dev desarrollo

For use in Claude.ai and ChatGPT

Download Skill