Marketplace MCP Skill
Claude skill for building and deploying MCP servers to FastMCP Cloud
Ask AI about Marketplace MCP Skill
Powered by Claude Β· Grounded in docs
I know everything about Marketplace MCP Skill. Ask me about installation, configuration, usage, or troubleshooting.
0/500
Reviews
Documentation
Marketplace MCP Skill
A Claude skill for building and deploying MCP servers to FastMCP Cloud.
Overview
This skill provides patterns, templates, and guidance for creating production-ready MCP servers with FastMCP 2 that can be deployed to the fastmcp.cloud marketplace.
Installation
Add this skill to your Claude Code configuration:
claude mcp add-skill marketplace-mcp-skill
Or add to your .claude/settings.json:
{
"skills": [
"github:7robots/marketplace-mcp-skill"
]
}
Usage
When building an MCP server, invoke the skill:
/marketplace-mcp
Or simply describe your task and Claude will use the skill automatically:
"Build an MCP server for the GitHub API and deploy it to fastmcp.cloud"
What's Included
Skill Guide (SKILL.md)
- Quick start workflow
- Development process
- Tool naming conventions
- Response format patterns
- Error handling patterns
- Secrets management
Reference Documentation
- FastMCP Patterns: Tool registration, Pydantic models, pagination, HTTP client patterns
- Deployment Guide: fastmcp.cloud deployment, secrets management, CI/CD
Example Server (example/)
Complete, deployable MCP server template demonstrating all best practices:
- 5 example tools with full documentation
- JSON and Markdown response formats
- Pagination support
- Error handling
- Ready for fastmcp.cloud deployment
Quick Start
1. Create a new MCP server
mkdir my-mcp-server && cd my-mcp-server
uv venv
uv pip install "fastmcp>=2.0.0,<3" httpx pydantic python-dotenv
2. Create server.py
import os
from fastmcp import FastMCP
mcp = FastMCP("my-service-mcp")
@mcp.tool()
async def my_tool(param: str) -> str:
"""Tool description."""
return f"Result: {param}"
if __name__ == "__main__":
host = os.getenv("MCP_HOST", "0.0.0.0")
port = int(os.getenv("MCP_PORT", "8000"))
mcp.run(transport="http", host=host, port=port)
3. Test locally
python server.py
# Or: fastmcp inspect server.py
4. Deploy to fastmcp.cloud
# Create GitHub repo
gh repo create my-mcp-server --public --source=. --push
# Then visit fastmcp.cloud to connect and deploy
Your server will be available at: https://my-mcp-server.fastmcp.app/mcp
Secrets Management
Local Development
Create .env file (never commit):
API_KEY=your-secret-key
DATABASE_URL=postgresql://user:pass@host/db
FastMCP Cloud
- Go to project settings on fastmcp.cloud
- Add environment variables under "Secrets"
- Variables are encrypted and injected at runtime
External Resources
- FastMCP Documentation: gofastmcp.com
- FastMCP Docs MCP Server:
https://gofastmcp.com/mcp - FastMCP Cloud: fastmcp.cloud
- MCP Protocol: modelcontextprotocol.io
File Structure
marketplace-mcp-skill/
βββ SKILL.md # Main skill file for Claude
βββ README.md # This file
βββ reference/
β βββ fastmcp_patterns.md # FastMCP 2 implementation patterns
β βββ deployment.md # Deployment and secrets guide
βββ example/
βββ server.py # Complete example server
βββ pyproject.toml # Dependencies
βββ .gitignore # Git ignore template
License
MIT
