Medium Accelerator

This MCP server lets downstream clients, like Claude Desktop, input a Medium archive URL and receive a compiled set of links to recent articles, a concise summary of those articles, and the ability to save and recall summaries locally for quick access.

How to use

You interact with this MCP server through an MCP client. Provide a Medium archive URL such as an article tag archive; the server will extract the article links, generate summaries (including Italian if you prefer), and store them in a local memory for fast retrieval later.

How to install

Prerequisites you need to have before starting:

• Python 3.10 or newer
• Claude Desktop

Step by step setup to run the MCP server locally:

# 1. Clone the project
git clone https://github.com/crtdaniele/mcp-medium-accelerator
cd mcp-medium-accelerator

# 2. Create and activate a virtual environment
python -m venv venv
source venv/bin/activate

# 3. Install Python dependencies
pip install -r requirements.txt

# 4. (Optional) Update requirements after adding new libraries
pip freeze > requirements.txt

How to run the MCP server

To run in development mode with hot reloading use the development command, otherwise start in normal mode.

mcp dev main.py

mcp run main.py

Claude Desktop installation

Install the MCP server into Claude Desktop or configure the client to connect to the local MCP server as described below.

mcp install main.py

{
  "mcpServers": {
    "mcp-medium-accelerator": {
      "command": "/opt/homebrew/bin/uv",
      "args": [
        "run",
        "--with",
        "bs4",
        "--with",
        "httpx",
        "--with",
        "datetime",
        "--with",
        "tinydb",
        "--with",
        "mcp[cli]",
        "mcp",
        "run",
        "/your-local-path/main.py"
      ]
    }
  }
}

Notes on usage

If you prefer a direct local command instead of the GUI path, you can run either the dev or normal command shown above. The dev mode provides hot reloading during development, while the normal mode runs the server without hot reload.

Available tools

extract_article_links

Extracts article links from a Medium archive URL and returns a list of article URLs.

extract_article_text

Extracts the content of a Medium article from its URL and returns the article text. It prompts you to save the summary via save_summary.

save_summary

Saves a summary of an article including the title, URL, and tags. Returns a status message.

list_summaries

Lists all saved summaries stored locally.