Claude-skill-registry kungfu

Use when finding new AI agent skills, discovering capabilities, installing skills from GitHub, searching skill marketplaces, or expanding what Claude can do - like Neo downloading martial arts in The Matrix

install

source · Clone the upstream repo

git clone https://github.com/majiayu000/claude-skill-registry

Claude Code · Install into ~/.claude/skills/

T=$(mktemp -d) && git clone --depth=1 https://github.com/majiayu000/claude-skill-registry "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/data/kungfu" ~/.claude/skills/majiayu000-claude-skill-registry-kungfu && rm -rf "$T"

manifest: skills/data/kungfu/SKILL.md

tags

#skill-discovery #github-search #agent-extensions #skill-installation #meta-skill

source content

<objective> Discover, evaluate, and install AI agent skills from the ecosystem. Like Neo downloading kung fu in The Matrix - expand capabilities by finding and integrating new skills. </objective> <discovery-strategy> Search GitHub for skill collections using: - Repos matching "awesome*skills" or "claude*skills" - Repos containing SKILL.md files (query: `path:SKILL.md claude`) - Topics: claude-code, ai-skills, agent-skills

Known curated collections (verify availability before citing):

VoltAgent/awesome-agent-skills
composioHQ/awesome-claude-code-skills
sickn33/antigravity-awesome-skills

If curated lists are unavailable, search GitHub directly. Prioritize repos updated within 6 months with meaningful star counts. </discovery-strategy>

<skill-format> Valid skills: SKILL.md with frontmatter (name, "Use when..." description, triggers). Optional scripts/tool for executables (require user approval before install). </skill-format> <quality-signals> **Evaluate:** GitHub stars, recent activity, SKILL.md quality, documentation, relevance to user's need.

Skip: No SKILL.md, abandoned (no commits in 1+ year AND unresponsive), <10 stars (unless from trusted source), duplicates already-installed skill. </quality-signals>

<security> If a skill includes executable scripts in `scripts/`, ALWAYS: 1. Show the user the script contents 2. Explain what the script does 3. Get explicit approval before installing

Never auto-execute downloaded scripts. </security>

<installation> Install to the project's `.claude/skills/<skill-name>/` directory. For global installation, use `~/.claude/skills/`.

Before installing:

Check for existing skill with same name
If conflict: ask user to overwrite, rename, or skip

After downloading:

Validate SKILL.md parses without YAML errors
Confirm required frontmatter (name, description, triggers) exists
Test with
```
/skill <name>
```
or a natural trigger phrase

If install fails: Remove any partially downloaded files and report the specific failure. </installation>

<workflows> <search>Query GitHub and curated lists, evaluate candidates against quality signals, present top 3-5 options with name, description, star count, last update, and install instructions.</search>

<install>Fetch skill, run security review if scripts present, validate format, install to skills directory, verify it loads.</install>

<audit>List installed skills from

.claude/skills/

, check source repos for updates, identify unused skills for removal.</audit> </workflows>