refactor-clean

Identify and remove dead code with test-backed verification.

182|28|Updated Nov 29, 2025
One-click install
npx skills add https://github.com/julianromli/ai-skills --skill refactor-clean-julianromli
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: refactor-clean
Source: https://github.com/julianromli/ai-skills/tree/main/skills/refactor-clean
Command: npx skills add https://github.com/julianromli/ai-skills --skill refactor-clean-julianromli

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Safely identify and remove dead code with test-backed verification to prevent regressions during refactors.

Core Features & Use Cases

  • Detect unused exports, dead utilities, and duplicate code across a codebase.
  • Recommend safe removals one at a time and verify with tests to guard against regressions.
  • Use during maintenance windows or before major refactors to prune tech debt without breaking behavior.

Quick Start

Establish a baseline, identify dead-code candidates, and remove one SAFE item at a time with immediate test verification.

Frequently Asked Questions about refactor-clean

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I safely remove dead code from a codebase without breaking existing behavior?

To safely remove dead code, establish a test baseline, categorize unused exports and stale utilities into risk levels, and delete one SAFE item at a time with immediate test re-verification to prevent regressions.

What is the best way to clean up unused exports and stale utilities before a major refactor?

The best way to clean up unused exports before a major refactor is to establish a test baseline, categorize dead-code candidates by risk, and incrementally remove SAFE items while verifying tests to guard against regressions.

How does test-backed verification work during dead code cleanup?

Test-backed verification works by enforcing a passing test baseline before deletion, then re-running tests immediately after removing one SAFE dead-code item to confirm no regressions occur.

When should I categorize dead code findings into SAFE, CAUTION, and DANGER levels?

You should categorize dead code findings into SAFE, CAUTION, and DANGER levels when pruning tech debt across modules and repositories, ensuring only SAFE items are deleted incrementally while higher-risk items require manual review.

Can I delete multiple dead code items at once during a maintenance window?

No, you should delete one SAFE dead-code item at a time. This incremental approach requires immediate test re-verification after each removal to ensure no unexpected regressions break the codebase.