skill-improve

Run a test-fix-retest loop with static checks to improve AI skills.

Updated May 24, 2026
One-click install
npx skills add https://github.com/LeKay/ai-game --skill skill-improve-lekay
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: skill-improve
Source: https://github.com/LeKay/ai-game/tree/main/.claude/skills/skill-improve
Command: npx skills add https://github.com/LeKay/ai-game --skill skill-improve-lekay

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill automates the process of improving a skill by running a test-fix-retest loop, including static checks, proposing fixes, retesting, and deciding whether to keep or revert changes.

Core Features & Use Cases

  • Test-Fix-Retest Loop: Runs static checks, proposes targeted fixes, rewrites the skill, re-tests, and keeps or reverts based on score change.
  • Static Checks: Identifies issues like missing frontmatter fields, incorrect phase counts, missing verdict keywords, missing ask-before-write language, and more.
  • Dynamic Category Checks: If a skill has a category assigned, it also runs category-specific checks.
  • User Interaction: Provides a clear step-by-step guide and allows users to accept or reject proposed changes.

Quick Start

Run the skill-improve command with the skill name as an argument, e.g., /skill-improve tech-debt.

Frequently Asked Questions about skill-improve

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

FAQPage Schema
How do I automate test-fix-retest loops for AI skill improvement?

Automate AI skill improvement by running a test-fix-retest loop that performs static checks, proposes targeted fixes, rewrites the skill, and retests to keep or revert changes based on score.

What static checks are performed during automated skill improvement?

Static checks during skill improvement identify missing frontmatter fields, incorrect phase counts, missing verdict keywords, and missing ask-before-write language, plus dynamic category-specific checks.

How do I run the skill-improve command to fix and test my AI skill?

Run the skill-improve command with the skill name as an argument, such as /skill-improve tech-debt, to trigger the automated test-fix-retest loop and view proposed changes.

Do I need access to source code to automate software testing and fixes?

Yes, automating software testing and fixes requires access to the skill's source code and the ability to execute tests to validate proposed changes before keeping or reverting them.

Can I reject proposed fixes during the automated test-fix-retest cycle?

Yes, you can reject proposed fixes during the test-fix-retest cycle, as the skill provides a step-by-step guide that allows users to accept or reject each proposed change.

What happens if a proposed code quality fix lowers the test score?

If a proposed code quality fix lowers the test score during the retest phase, the automated loop decides whether to revert the changes to maintain the previous state.