MCP Server for OneNote Integration

The Model Context Protocol (MCP) server enables AI assistants to interact with Microsoft OneNote for reading, creating, and managing notebooks, sections, and pages.

Details

GitHub Repositorygithub.com/rajvirtual/MCP-Servers

GitHub Stars0

Language

TypeScript

Description

The MCP server provides seamless integration capabilities between AI models and Microsoft OneNote. Built on Node.js and TypeScript, the server uses the Microsoft Graph API for communication and MSAL for authentication. It offers tools for note-taking automation, content creation, and HTML-to-text conversions, compatible with Docker for containerized deployment. MCP unlocks the potential of AI-driven workflows within productivity applications.

Features

Authentication via MSAL with device code flow
Tools for reading and creating content in OneNote
Token caching for streamlined authentication
Docker compatibility for deployment
Integration with Microsoft Graph API for communication
Built-in support for RAG operations with HTML-to-text conversion

Running Instructions

Installation

Clone the repository.

    git clone https://github.com/rajvirtual/MCP-Servers/tree/master/onenote

Install dependencies:

    npm install

Create a .env.local file with your Azure client ID:

    CLIENT_ID=your-client-id-from-azure

Development

Run the application in development mode:

    npm run dev

Building

Compile TypeScript to JavaScript:

    npm run build

Running

Run the compiled application:

    npm start

Docker Instructions

Create a data directory for persistence:

    mkdir -p data

Build the Docker image:

    docker build -t onenote-mcp-server .

Run the container:

    docker run -d \
      --name onenote-mcp-server \
      -e CLIENT_ID=your-client-id \
      -v $(pwd)/data:/app/dist \
      onenote-mcp-server

Authentication:

Check the device code file within the container:

    cat data/device-code.txt

Visit the given URL and enter the code to authenticate. Tokens are cached in data/token-cache.json for future use.

Tools

MCP Tools

- **onenote-read**: Reads content from notebooks, sections, or pages. Parameters include `pageId`, `sectionId`, `notebookId`, `includeContent`, and `includeMetadata`. - **onenote-create**: Creates new content, including notebooks, sections, or pages. Parameters include `type`, `title`, `content`, and `parentId`.

Use Cases

Automating note-taking workflows
Enabling AI-driven search and retrieval operations within OneNote
Contextual assistance for summarization and editing of OneNote content
Creating structured notes programmatically
Integrating AI for content generation within OneNote notebooks

Frequently Asked Questions

No frequently asked questions available.