Google新闻解析服务

Google News Parser (FastMCP Server)

Status: Live & Tested
Protocol: FastMCP
Authentication: None

This project provides a simple, lightweight FastMCP (Model Context Protocol) server that parses the Google News RSS feed. It exposes a single tool, get_google_news_feed, to fetch the latest headlines, descriptions, and source metadata.

The server is built using the fastmcp library for the MCP implementation and feedparser for parsing the RSS feed.

🚀 Live Server Endpoint

The MCP server is deployed on FastMCP Cloud and is publicly accessible for immediate use by any MCP-compatible client or AI model.

• MCP Endpoint URL: https://google-news-parser.fastmcp.app/mcp

✅ Server Status & Testing

The endpoint is live and has been fully tested and confirmed working with the following clients and inspectors:

• Cursor.ai (via mcp.json context)
• @modelcontextprotocol/inspector
• FastMCP's built-in inspector

🛠️ Tool Definition

The server exposes the following tool:

get_google_news_feed

• Description: Get the latest headlines and metadata from the Google News RSS feed.
• Tags: google news, news feed, rss, headlines
• Arguments:
  • max_results (int, optional): The maximum number of articles to return. Default: 3.

---

💻 Example Usage

You can interact with this MCP server using any standard HTTP client or a compatible editor like Cursor.ai.

Using Cursor.ai

To use this tool directly within the Cursor.ai editor, follow these two steps:

1. Create mcp.json: In the root of your project folder, create a file named mcp.json and add the URL of the server's manifest.

   mcp.json

   复制

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

2. Prompt the Chat: Cursor will automatically detect the file and load the tool. You can now just ask the AI to use it in plain English.

   User Prompt:

   Hey, can you get me the top 2 headlines from Google News using the get_google_news_feed tool?

Cursor will then execute the tool against your live server and return the results directly in the chat. print(f"\nAn error occurred: {e}")