generic-mcp: Model Context Protocol Server Template

A robust, extensible Model Context Protocol (MCP) server template for building tool-based automation APIs. Easily add, expose, and document any set of tools (file, database, codegen, etc.) for LLM agents and automation workflows.

Features

• Easily add and expose any tool (file, database, codegen, etc.) as an MCP tool
• RESTful tool interface for LLM agents and automation
• OpenAPI schema available at /mcp/openapi.json
• Built with Python, FastMCP, Pydantic, and Uvicorn

Project Structure

generic-mcp/
├── mcp_server.py         # Main MCP server entrypoint (rename as needed)
├── README.md             # This file
├── pyproject.toml        # Python project config
└── ...

Setup

1. Install dependencies:

   uv pip install -r requirements.txt

2. Run the server:

   uv run python file_mcp.py --root <workspace-root>
   # Example:
   uv run python file_mcp.py --root C:\Users\yourname\ai-play-ground

Usage

• The server exposes your custom tools via HTTP at http://localhost:8123/mcp.
• OpenAPI schema: http://localhost:8123/mcp/openapi.json
• Use an MCP-compatible agent or client to call your defined tools.

Tool API

Define your own tools using the @mcp.tool() decorator in your main server file. Each tool will be automatically exposed via HTTP and documented in the OpenAPI schema.

Example:

@mcp.tool()
def my_tool(param1: str, param2: int) -> dict:
	"""Describe what this tool does."""
	# ...your logic...
	return {"result": ...}

See the OpenAPI schema for details on all available tools and their parameters.

Best Practices

• Keep all tool logic in your main server file or organize by feature as needed.
• Use the @mcp.tool() decorator for every tool you want to expose.
• Log important events and errors using the built-in logger.

License

MIT