Blog RSS MCP Server

A small FastMCP-based service exposing two tools for working with a blog RSS feed:

Blog RSS MCP Server

A minimal Model Context Protocol (MCP) server that exposes tools for working with a blog’s RSS/Atom feed. It lists posts, fetches post content, returns recent posts, basic blog info, and can search post content on-demand.

What’s in here

• main_server.py — MCP server with tools:

  • list_blog_posts() → [{title, slug (url), pubDate}]

  • get_blog_post(slug) → {slug, url, content} (content extracted from page HTML, preferring )

  • get_recent_posts(count=5) → recent posts (simple reverse order)

  • get_blog_info() → blog metadata

  • search_full_text(query) → on-demand search (fetches each post and searches the text directly)

• run_local_tests.py — quick local smoke test that exercises the tools without starting the MCP server.

• Dockerfile — container image for running the server.

• requirements.txt — Python dependencies.

Requirements

• Python 3.10+

• Recommended: a virtual environment (e.g., .venv)

Install dependencies:

Note for Python 3.13+: feedparser imports the removed stdlib module cgi; this repo includes python-legacy-cgi in requirements to ensure compatibility.

Configuration

Set environment variables as needed:

• RSS_FEED_URL (required) → Full URL to your RSS/Atom feed (e.g., https://example.com/feed.xml)

• MCP_TRANSPORT (optional) → stdio (default), sse, or streamable-http

• MCP_MOUNT_PATH (optional) → mount path for applicable transports

Example:

Run locally (MCP server)

With a venv:

Local smoke test (no MCP client required)

That script will print blog info, list posts, fetch one post’s content, and run a simple on-demand search.

Docker

Build the image:

Run with stdio transport (for MCP clients that spawn the container and talk over stdio):

Run with basic HTTP transport (optional):

Notes & troubleshooting

• Logs go to stderr; stdio transport messages go to stdout (safe for MCP clients).

• If HTTPS feed fetches fail in a minimal base image, install system CA certificates in your container.

• If mcp.run() signature changes between versions, inspect FastMCP.run in your installed package and adjust the call accordingly.

Example:

Running locally

Start the server directly with Python (development):

If MCP_TRANSPORT=stdio, the process runs attached to STDIN/STDOUT so a supervising client can manage lifecycle and communicate over stdio.

Docker

Build the image locally:

Run attached to your terminal (stdio transport):

Or run detached with an HTTP/streamable transport (adjust ports as needed):

Search

• search_full_text(query) performs a simple case-insensitive substring search on-demand. It fetches each post’s content live and extracts plain text, so results always reflect the current site content.

Custom catalog and client integration

• The repo includes custom_catalog_for_blog_mcp.txt (example YAML you can copy into a client catalog). Use MCP_TRANSPORT=stdio when you want the client to spawn the container and communicate over stdio.

Troubleshooting

• If you see import errors for feedparser or beautifulsoup4, run pip install -r requirements.txt in the same environment.

• If posts are missing from list_blog_posts(), verify the feed provides published or pubDate; the server is defensive but you can inspect the raw parsed feed:

• If mcp.run() fails due to signature mismatch, inspect mcp.server.fastmcp.FastMCP.run in your installed package and adjust the call in main_server.py.

Development & tests

• Add pytest tests that mock feedparser.parse to exercise list_blog_posts, get_blog_post, and search_full_text.

• Consider adding a GitHub Actions workflow to run tests on push.

Security & production notes

• Do not expose the server publicly without authentication or a reverse proxy with TLS.

• When building/persisting indexes, secure the storage and limit who can trigger a rebuild.

Next steps I can help with

• Wire _build_fulltext_index() to run at startup when BUILD_FULLTEXT_INDEX is enabled (optionally persist to disk).

• Add unit tests and a CI workflow.

• Add a HEALTHCHECK to the Dockerfile and/or a small HTTP health endpoint.

Tell me which of those you'd like and I'll implement it.