Fourth Brain

This MCP server provides a Notion-backed knowledge base for Fourth’s marketing content, enabling Claude.ai to securely query curated information through the Model Context Protocol. It combines a locally hosted Python server with a Notion-backed data store, and offers a straightforward path to deployment and testing with MCP tooling.

How to use

You will connect an MCP client to the Fourth Brain MCP server to query and retrieve marketing content. Start by running the local server, then test usage with the MCP Inspector, and finally deploy to a hosted URL for Claude.ai integration.

How to install

Prerequisites you need before starting:

Python 3.8 or newer must be installed on your machine.

Git is required to clone and push your project to a hosting service.

Follow these concrete steps to set up the server locally and prepare your environment.

pip install -r requirements.txt
python populate_notion.py --parent-page <page_id>

After population completes, run the local server and verify it responds at the designated MCP endpoint.

python server.py

Test the MCP setup with the MCP Inspector to ensure the server announces its MCP interface.

fastmcp dev server.py

If you plan to deploy to Railway, push your repository to GitHub and set environment variables in the Railway dashboard. The deployment will auto-run from GitHub.

# Push to GitHub
git remote add origin https://github.com/Rev4nchist/fourth-brain-demo.git
git push -u origin main

After auto-deploy, configure the MCP URL in Claude.ai as described in integrations. The typical URL will look like the hosted app’s /mcp path.

Additional setup for Claude.ai integration

In Claude.ai, go to Settings > Integrations. Add a Custom Connector and enter the hosted MCP URL, for example: https://your-app.up.railway.app/mcp. Create a Project and upload your skills files from the skills folder.

Available tools

search_knowledge

Search across all knowledge databases to locate relevant documents and content pieces.

browse_library

Navigate the knowledge structure by area or domain to understand content organization.

get_document

Retrieve full content of a specific page or document from the Notion-backed store.

ask_question

Perform RAG-style Q&A with confidence scoring to answer questions using the knowledge base.

list_content_areas

List all knowledge domains available in the system.