🔍 Skills Search MCP

<div align="center">

A powerful MCP server that brings the skills.sh ecosystem directly to your AI agents

</div>

---

✨ Why Skills Search MCP?

"The best AI agents are those with the right skills."

Finding and installing the right skills for your AI agents shouldn't be a manual research project. Skills Search MCP bridges your AI agents directly to the open-source skills ecosystem, making it effortless to discover, install, and manage skills—all from within your agent workflow.

---

🚀 Features

• 🔎 Smart Discovery — Search skills by technologies (React, Python, Docker, etc.) or task domains (testing, deployment, linting)
• 📦 Full Management — Install, uninstall, and update skills without leaving your agent
• ⚡ Parallel Search — Lightning-fast async searches across multiple technologies
• 🎯 Type-Safe — Built with Pydantic for robust input validation
• 🧹 Clean Output — Automatically strips ANSI codes for readable responses
• 🔌 Universal CLI Access — Run any skills command directly when needed

---

📋 Available Tools

Tool	Description
find_useful_skills	Discover skills by technologies and task domain
list_installed_skills	See what skills are already installed
get_skill_info	Get details about a specific skill
install_skill	Install a skill locally
uninstall_skill	Remove a skill
update_skills	Update installed skills to latest versions
run_skill_command	Execute raw skills CLI commands

---

🛠️ Installation

Prerequisites

• Python 3.10+
• Node.js & npm (for the skills CLI)

Install Dependencies

pip install fastmcp pydantic

Or with uv:

uv pip install fastmcp pydantic

Install the Skills CLI (if you haven't)

npm install -g @modelcontextprotocol/skills

---

⚡ Quick Start

1. Run the Server

PYTHONPATH=src python3 src/skills_search_mcp/server.py

2. Connect to Claude Desktop (macOS)

Add to ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "skills-search": {
      "command": "python3",
      "args": ["-P", "/path/to/skills-mcp/src", "src/skills_search_mcp/server.py"]
    }
  }
}

3. Connect to Cursor

Add to Cursor settings (or cursor.mcp.json):

{
  "mcpServers": {
    "skills-search": {
      "command": "python3",
      "args": ["-P", "/path/to/skills-mcp/src", "src/skills_search_mcp/server.py"]
    }
  }
}

---

💡 Usage Examples

Find React Skills

find_useful_skills(technologies=["react"])

Find Testing Skills for Python

find_useful_skills(technologies=["python"], task_domain="testing")

List Installed Skills

list_installed_skills()

Install a Skill

install_skill(skill_url="vercel-labs/agent-skills@vercel-react-best-practices")

Check for Updates

update_skills(update_all=True)

---

🔧 Configuration

Environment Variables

Variable	Description	Default
SKILLS_TIMEOUT	Timeout for skills commands (seconds)	60

---

🤝 Contributing

Contributions are welcome! Here's how you can help:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

---

📚 Resources

• skills.sh — Browse the skills ecosystem
• MCP Documentation — Learn about MCP
• FastMCP — Python SDK

---

⭐ Show Your Support

If this project helps you build better AI agents, give it a ⭐️!

---

<div align="center">

Built with ❤️ for the AI agent community

</div>