📡 RSS MCP Server

📡 RSS MCP Server

A modular Model Context Protocol (MCP) server for tracking, filtering, and surfacing PNT/GNSS news from custom RSS feeds with natural language interface via Claude Desktop.

✨ Features

• 🔗 Native MCP Integration - Direct integration with Claude Desktop
• 📰 RSS Feed Processing - Fetches and parses multiple RSS feeds
• 🏷️ Tag-based Filtering - Content filtering using configurable tags
• 💬 Natural Language Interface - AI-powered RSS management
• 📊 State Management - Persistent read/unread article tracking
• 🔍 Search Capabilities - Keyword and tag-based article search
• 🛡️ Error Handling - Graceful fallbacks and error recovery

🚀 Quick Start

Prerequisites

• Python 3.10+
• Claude Desktop
• Virtual environment (recommended)

Installation

# Clone the repository
cd rss-mcp-server

# Create and activate virtual environment
python3 -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install dependencies
pip install mcp feedparser pyyaml httpx

Configuration

1. Create feeds.yaml (if not exists):

- url: https://insidegnss.com/feed/
  tags: [C-band, TrustPoint, GPSIA, LEO, PNT]
- url: https://spacenews.com/feed/
  tags: [LEO, spoofing, spectrum]
- url: https://www.gpsworld.com/feed/
  tags: [GNSS, spoof, jamming, M-code]

2. Configure Claude Desktop:

Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS):

{
  "mcpServers": {
    "rss-mcp-server": {
      "command": "/path/to/your/project/.venv/bin/python",
      "args": ["/path/to/your/project/main.py"],
      "cwd": "/path/to/your/project"
    }
  }
}

⚠️ Replace paths with your actual project location

3. Restart Claude Desktop completely (quit and reopen)

Testing

# Test server locally
python main.py
# Should output: "RSS MCP Server initialized successfully"

🤖 Usage

Once integrated with Claude Desktop, use natural language commands:

"Update my RSS feeds"
"Show me articles about GPS jamming"
"Search for LEO satellite articles"
"Mark the first 5 articles as read"
"Find articles about spoofing"
"Show me 10 recent unread articles"

📡 MCP Tools

Tool	Description	Parameters
update_rss_feeds	Fetch latest articles from all feeds	None
get_unread_articles	Get unread articles	limit (optional)
mark_articles_read	Mark articles as read	article_ids (optional)
search_articles	Search articles by keywords	query, tags (optional)

📊 MCP Resources

Resource URI	Description
rss://feeds/unread	All unread articles (JSON)
rss://feeds/all	All articles (JSON)

📁 Project Structure

rss-mcp-server/
├── main.py              # MCP server implementation
├── processor.py         # RSS feed processing logic
├── state.py            # State management for read/unread tracking
├── feeds.yaml          # Feed configuration (direct list format)
├── state.json          # Runtime state (auto-created)
├── requirements.txt    # Python dependencies
├── README.md          # This file
└── .venv/             # Virtual environment

⚙️ Configuration Details

feeds.yaml Format

✅ Correct format (direct list):

- url: https://example.com/feed/
  tags: [tag1, tag2, tag3]
- url: https://another.com/rss/
  tags: [tag4, tag5]

❌ Incorrect format (with wrapper):

feeds:  # Don't use this wrapper
  - url: https://example.com/feed/

Tags

Tags are used for:

• Content filtering during search
• Categorizing articles by topic
• Enhancing search relevance

Example tags for PNT/GNSS content:

• GPS, GNSS, PNT
• jamming, spoofing, interference
• LEO, MEO, satellites
• military, defense, civilian
• M-code, anti-jam, resilient

🔧 Troubleshooting

Common Issues

Server not appearing in Claude Desktop:

• Verify claude_desktop_config.json syntax and paths
• Ensure all paths are absolute
• Restart Claude Desktop completely

Import errors:

pip install mcp feedparser pyyaml httpx

feeds.yaml format errors:

• Use direct list format (start with -)
• No feeds: wrapper needed
• Ensure proper YAML syntax

File permission errors:

• Ensure write permissions for state.json
• Check directory permissions

Debug Steps

1. Test locally: python main.py
2. Check dependencies: pip list | grep -E "(mcp|feedparser|pyyaml|httpx)"
3. Validate config: Check Claude Desktop logs
4. Verify feeds: Test feed URLs in browser

🧪 Development

Running Tests

# Test RSS feed processing
python -c "from processor import RSSProcessor; import asyncio; asyncio.run(RSSProcessor().initialize())"

# Test state management
python -c "from state import StateManager; sm = StateManager(); print('State manager working')"

Adding New Feeds

1. Add feed URL to feeds.yaml
2. Configure appropriate tags
3. Restart the MCP server
4. Update feeds: "Update my RSS feeds"

Debugging

Enable debug output by modifying main.py:

import sys
print("Debug info", file=sys.stderr)

📈 Performance

• Feed Updates: ~2-5 seconds for 3-5 feeds
• Article Search: <1 second for typical queries
• State Persistence: Automatic on read/unread changes
• Memory Usage: ~10-20MB typical operation

🤝 Contributing

1. Fork the repository
2. Create a feature branch
3. Make changes with tests
4. Submit a pull request

📄 License

MIT License - see LICENSE file for details

🆘 Support

• Issues: Create GitHub issue with logs
• Questions: Tag @anil.goparaju
• Documentation: See Google Drive docs

🎯 Use Cases

Daily News Briefing:

"Update feeds and show me unread articles about GPS jamming and spoofing"

Research Assistance:

"Search for articles about LEO PNT systems published this week"

Competitive Intelligence:

"Find articles mentioning TrustPoint or military navigation systems"

Content Curation:

"Show me all unread articles, then mark the first 10 as read"

📊 Sample Output

When you search for "GPS jamming", you get structured results:

{
  "count": 3,
  "articles": [
    {
      "id": "251bc714...",
      "title": "U.S. Army Taking a Layered Approach to PNT",
      "link": "https://insidegnss.com/...",
      "published": "2025-06-03T16:58:20",
      "author": "Renee Knight",
      "feed_tags": ["C-band", "TrustPoint", "GPSIA", "LEO", "PNT"],
      "content": "...anti-jamming capabilities..."
    }
  ]
}

---

Last Updated: June 5, 2025
Status: ✅ Production Ready