Official Substack / Medium MCP Server

What is Official Substack / Medium MCP Server?

Official Substack / Medium MCP Server is an open-source Model Context Protocol (MCP) implementation that connects Claude to your Substack and Medium writing. It allows Claude to access, cache, and semantically search your published blog posts, providing personalized writing assistance.

How to use Official Substack / Medium MCP Server?

Clone the repository, set up a Python 3.10+ environment with uv or pip, and configure your Substack/Medium RSS URLs in a config.json file. Then connect to Claude Desktop by adding the server to the claude_desktop_config.json configuration. Once running, Claude can access individual essay resources and use the search_writing and refresh_content tools.

Key features of Official Substack / Medium MCP Server

• Retrieves and permanently caches blog posts from Substack and Medium
• Uses embeddings to find the most relevant essays via semantic search
• Makes individual essays available as separate resources for Claude
• Preloads all content and generates embeddings at startup
• Supports configurable caching duration and maximum posts to fetch
• Provides two tools: search_writing and refresh_content

Use cases of Official Substack / Medium MCP Server

• Ask Claude to find essays discussing a specific topic across your Substack and Medium archives
• Reference individual essays by title when requesting writing feedback or idea development
• Automatically refresh your cached content to keep Claude up to date with new posts

FAQ from Official Substack / Medium MCP Server

What platforms does Official Substack / Medium MCP Server support?

Substack and Medium are supported via their RSS feeds.

How does caching work in Official Substack / Medium MCP Server?

Content is stored on disk and persists between sessions. The cache duration is configurable (default 10080 minutes / 1 week). Content and embeddings are automatically refreshed at startup and on demand using the refresh_content tool.

What are the prerequisites for using Official Substack / Medium MCP Server?

Python 3.10 or higher, the latest Claude Desktop, and a Substack or Medium account with published public content.

How can I troubleshoot if the tool doesn't appear in Claude Desktop?

Check that your Claude Desktop configuration file has correct absolute paths, verify all packages are installed, and restart Claude Desktop. You can also try the shell script alternative for the uv command.

How does semantic search work in Official Substack / Medium MCP Server?

Each essay is converted into embeddings at startup. When you use the search_writing tool, embeddings are compared to find the most relevant essays, with the number of results configurable via similar_posts_count.