Quickstart Guide to Building an MCP Server in Python

- Uses the official MCP Python SDK.
- Provides a calculator tool (add) and a dynamic greeting resource.
- Employs FastMCP for server initialization and tool/resource decorators.
- Demonstrates testing with the MCP Inspector via mcp dev.
- Shows client integration with Claude Desktop via mcp install.
- Covers deployment topics: security, scalability, and monitoring.

To integrate MCP into your Python project, it's recommended to use uv, a Python package manager:

uv add "mcp[cli]"

Alternatively, if you're using pip:

pip install mcp

Let's create a simple MCP server that offers a calculator tool and a personalized greeting resource:

```python
from mcp.server.fastmcp import FastMCP

When deploying your MCP server:

- Security: Implement authentication mechanisms, such as API keys or OAuth, to control access.
- Scalability: Utilize containerization tools like Docker to manage deployments across various environments.
- Monitoring: Set up logging and monitoring to track server performance and diagnose issues promptly.

@mcp.tool()
def add(a: int, b: int) -> int:
"""Adds two numbers."""
return a + b