remarkable-mcp 0.1.0

MCP server for accessing reMarkable tablet data

reMarkable MCP Server

An MCP (Model Context Protocol) server for accessing your reMarkable tablet data through the reMarkable Cloud.

Quick Install

What It Does

• Read typed text directly from notebooks (v3+ software, no OCR needed)
• Browse and search your document library
• Access recent files with content previews
• OCR for handwritten content via pytesseract
• MCP resources and prompts for deeper integration

Installation

Using uvx (Recommended)

# Get your reMarkable token
uvx remarkable-mcp --register YOUR_ONE_TIME_CODE

Click the Quick Install badges above, or configure manually.

From Source

git clone https://github.com/SamMorrowDrums/remarkable-mcp.git
cd remarkable-mcp
uv sync
uv run python server.py --register YOUR_ONE_TIME_CODE

Setup

1. Get a One-Time Code

Go to my.remarkable.com/device/browser/connect and generate a code.

2. Convert to Token

uvx remarkable-mcp --register YOUR_CODE

3. Configure MCP

VS Code — Add to .vscode/mcp.json:

{
  "inputs": [
    {
      "type": "promptString",
      "id": "remarkable-token",
      "description": "reMarkable API Token",
      "password": true
    }
  ],
  "servers": {
    "remarkable": {
      "command": "uvx",
      "args": ["remarkable-mcp"],
      "env": {
        "REMARKABLE_TOKEN": "${input:remarkable-token}"
      }
    }
  }
}

Your token is stored securely using VS Code's input system with password: true.

Claude Desktop — Add to claude_desktop_config.json:

{
  "mcpServers": {
    "remarkable": {
      "command": "uvx",
      "args": ["remarkable-mcp"],
      "env": {
        "REMARKABLE_TOKEN": "your-token-from-step-2"
      }
    }
  }
}

Tools

Tool	Description
remarkable_read	Extract text from a document
remarkable_browse	List files or search by name
remarkable_recent	Get recently modified documents
remarkable_status	Check connection status

All tools are read-only and return structured JSON with hints for next actions.

Resources

Recent documents are automatically registered as MCP resources on startup (if authenticated). Each document becomes available at remarkable://doc/{name}.

URI	Description
remarkable://doc/{name}	Content of a recent document
remarkable://folders	Complete folder hierarchy

Prompts

summarize_recent · find_notes · daily_review · export_document · organize_library · meeting_notes

Usage

remarkable_read("Meeting Notes - Nov 2025")
remarkable_browse("/")
remarkable_browse(query="meeting")
remarkable_recent(limit=5, include_preview=True)

Text Extraction

Typed text from v3+ notebooks is extracted natively via rmscene — no OCR required.

Handwritten content uses pytesseract for OCR. Make sure Tesseract is installed on your system:

# macOS
brew install tesseract

# Ubuntu/Debian
sudo apt-get install tesseract-ocr

# Arch
sudo pacman -S tesseract

PDF highlights and annotations are also extracted.

Design

Intent-based tools that map to what you actually want to do. Responses include hints for logical next steps. Errors explain what went wrong and how to fix it. Four tools cover most use cases.

Authentication

Set REMARKABLE_TOKEN in your MCP config (recommended), or the server will fall back to ~/.rmapi.

Development

uv sync --all-extras
uv run pytest test_server.py -v
uv run ruff check .
uv run ruff format .

Project Structure

remarkable-mcp/
├── server.py           # Entry point
├── remarkable_mcp/     # Main package
│   ├── server.py       # FastMCP server
│   ├── api.py          # Cloud API helpers
│   ├── extract.py      # Text extraction
│   ├── tools.py        # MCP tools
│   ├── resources.py    # MCP resources
│   └── prompts.py      # MCP prompts
└── test_server.py      # Tests

License

MIT

---

Built with rmapy, rmscene, and inspiration from Scrybble.