agentleFS
kinged007kinged007/agent-skill-search/SKILL.md

skill-search

Skill0 starsChanged 28 days agoMITView on GitHub

Universal skill search tool. Use grep like search terms to find relevant skills for your task. Always use this tool.

  • Installs packages

4 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions4 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 4

    1. 0oooooooo0/skillless★ 38same name
    2. taxueseek/session-digger★ 34same name
    3. AI-ModCon/genesis-skills★ 30same name
    4. ong6/skills★ 0same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repositoryOne other file it gives its agents.

The file

Files
SKILL.md

Skill Search

A portable tool that lets agents search through a catalog of SKILL.md files and load only the relevant ones — instead of dumping 100+ skills into context.

How It Works

The tool has two modes:

  1. MCP Server — for Claude Code, Claude Desktop, Codex, Gemini CLI, OpenClaw, Hermes, pi, or any MCP-capable agent. Exposes skill_search, skill_view, and skill_list as tools.
  2. CLI — for any agent. Run skill-search search <query> to find skills, skill-search view <name> to load one.

Setup

Install the package first (distribution name is agent-skill-search):

uv tool install agent-skill-search
# or: pip install agent-skill-search

Option A: MCP Server (recommended)

One command per client:

skill-search install claude-code   # or codex / gemini / openclaw / hermes / pi / claude-desktop

Manual fallback for any mcpServers-style JSON config:

{
  "mcpServers": {
    "skill-search": {
      "command": "/absolute/path/to/skill-search-mcp",
      "args": [],
      "env": {
        "SKILL_CATALOG_DIRS": "/path/to/your/skills:/path/to/more/skills"
      }
    }
  }
}

Or with uvx (note the --from, since the distribution is agent-skill-search):

{
  "mcpServers": {
    "skill-search": {
      "command": "uvx",
      "args": ["--from", "agent-skill-search", "skill-search-mcp"],
      "env": {
        "SKILL_CATALOG_DIRS": "~/.hermes/skills:~/.agents/skills"
      }
    }
  }
}

Full per-client instructions, the ChatGPT remote-HTTP setup, and troubleshooting: INSTALL.md.

Option B: CLI (any agent)

pip install agent-skill-search

# Set your catalog directories (colon-separated)
export SKILL_CATALOG_DIRS="$HOME/.hermes/skills:$HOME/.agents/skills"

# Search (space = AND, `|` = OR, quotes = phrase)
skill-search search "react performance"
skill-search search "planning|prd|guide"

# View a skill
skill-search view "vercel-react-best-practices"

# List all
skill-search list

Option C: Local development

cd skill-search
pip install -e .
export SKILL_CATALOG_DIRS="."
skill-search list

Catalog management

skill-search add <dir> copies skill directories from the source tree into the default catalog (~/.agents/skills-catalog) — the source is left untouched.

Search the catalog:

skill-search search "react performance"   # find skills
skill-search search "planning|prd|guide"  # OR search
skill-search "agent orchestration instructions"  # shorthand (== search)
skill-search view "react-perf"             # load one
skill-search --version                   # show version
skill-search update                      # upgrade to the latest release

Usage Pattern for Agents

  1. Discover — run skill_search(query="your topic") to find matching skills
  2. Evaluate — read the name + description from results
  3. Load — run skill_view(name="the-skill") to get full instructions
  4. Execute — follow the skill's procedure

This keeps context lean: only the catalog index (~3k tokens) plus the one skill you actually need.

Catalog Structure

Point the tool at directories containing SKILL.md files:

your-catalog/
├── react-performance/
│   └── SKILL.md
├── kubernetes-deploy/
│   └── SKILL.md
└── api-design/
    └── SKILL.md

Any directory tree containing SKILL.md files works — the tool scans recursively.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.