MCP Servers Using Python

FastMCP & FastAPI-MCP Examples

This repository is a collection of sample projects and practice files demonstrating how to build Model Context Protocol (MCP) servers using both the fastmcp (pure) and fastapi_mcp (FastAPI integration) libraries.

It includes examples for:

• Wrapping a pure function (calculator.py)
• Wrapping an external API (weather.py)
• Parsing an RSS feed (GoogleFeedParser.py)
• Auto-generating tools from a full FastAPI API (calculator-api.py)

🚀 Live Deployed Server: Google News Parser

The code from GoogleFeedParser.py (also organized in the MCP Deployment using FastMCP Cloud folder) is deployed as a live, public MCP server.

• Live MCP Endpoint: https://google-news-parser.fastmcp.app/mcp
• Deployment-Specific Repo: More information about this specific deployment can be found at:

Quick Usage

1. Using Cursor.ai

Create an mcp.json file in your project's root:

      {
          "mcpServers": {
              "google-news-parse": {
                  "url": "https://google-news-parser.fastmcp.app/mcp"
              }
          }
      }