Skip to content

KKKKhazix/Khazix-Skills

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Khazix-Skills

A collection of AI Skills for managing and evolving your skill library. These tools help you create, maintain, and continuously improve AI skills from GitHub repositories.

Skills Overview

Skill Description
github-to-skills Convert GitHub repos into AI skills automatically
skill-manager Manage skill lifecycle - check updates, list, delete
skill-evolution-manager Evolve skills based on user feedback and experience

github-to-skills

Automated factory for converting GitHub repositories into specialized AI skills.

Features

  • Fetches repository metadata (README, latest commit hash)
  • Creates standardized skill directory structure
  • Generates SKILL.md with extended frontmatter for lifecycle management
  • Creates wrapper scripts for tool invocation

Usage

/github-to-skills <github_url>

Or: "Package this repo into a skill: "

Example

/github-to-skills https://github.com/yt-dlp/yt-dlp

skill-manager

Lifecycle manager for GitHub-based skills.

Features

  • Audit: Scan local skills folder for GitHub-based skills
  • Check: Compare local commit hashes against remote HEAD
  • Report: Generate status report (Stale/Current)
  • Update: Guided workflow for upgrading skills
  • Inventory: List all skills, delete unwanted ones

Usage

/skill-manager check     # Scan for updates
/skill-manager list      # List all skills
/skill-manager delete <name>  # Remove a skill

Scripts

Script Purpose
scan_and_check.py Scan directories, parse frontmatter, check remote versions
update_helper.py Backup files before update
list_skills.py List installed skills with metadata
delete_skill.py Permanently remove a skill

skill-evolution-manager

Continuously improve skills based on user feedback and conversation insights.

Core Concepts

  1. Session Review: Analyze skill performance after conversations
  2. Experience Extraction: Convert feedback into structured evolution.json
  3. Smart Stitching: Persist learned best practices into SKILL.md

Usage

/evolve

Or: "Save this experience to the skill"

Workflow

  1. Review: Agent analyzes what worked/didn't work
  2. Extract: Creates structured JSON with preferences, fixes, custom prompts
  3. Persist: Merges into evolution.json
  4. Stitch: Updates SKILL.md with learned best practices

Scripts

Script Purpose
merge_evolution.py Incrementally merge new experience data
smart_stitch.py Generate/update best practices section in SKILL.md
align_all.py Batch re-stitch all skills after updates

Installation

  1. Clone this repository:
git clone https://github.com/KKKKhazix/Khazix-Skills.git
  1. Copy desired skills to your Claude skills directory:
# Windows
copy /E Khazix-Skills\github-to-skills %USERPROFILE%\.claude\skills\

# macOS/Linux
cp -r Khazix-Skills/github-to-skills ~/.claude/skills/
  1. Restart Claude to load the new skills.

Requirements

  • Python 3.8+
  • Git (for checking remote repositories)
  • PyYAML (pip install pyyaml)

How It Works

+------------------+     +----------------+     +------------------------+
| github-to-skills | --> | skill-manager  | --> | skill-evolution-manager|
+------------------+     +----------------+     +------------------------+
        |                       |                         |
    Create new             Maintain &                 Evolve &
    skills from            update skills              improve based
    GitHub repos                                      on feedback

The Complete Skill Lifecycle:

  1. Create: Use github-to-skills to wrap a GitHub repo as a skill
  2. Maintain: Use skill-manager to check for updates and upgrade
  3. Evolve: Use skill-evolution-manager to capture learnings and improve

License

MIT


Contributing

Contributions are welcome! Feel free to:

  • Report issues
  • Submit pull requests
  • Share your own skills

Author

KKKKhazix

If you find these skills useful, consider giving this repo a star!

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages