remarkable-mcp

An MCP server that gives Claude direct access to your reMarkable tablet's notebooks. Search documents, render handwritten pages to PDF, and let Claude transcribe your handwriting or convert hand-drawn diagrams into editable formats — all from your local machine, no API keys required.

How it works

The reMarkable desktop app syncs your notebooks to a local cache. This MCP server reads that cache and exposes it as tools that Claude can use directly.

reMarkable tablet
    → reMarkable desktop app (cloud sync)
    → Local cache (~/.../remarkable/desktop)
    → remarkable-mcp (rmc → SVG → cairosvg → PDF)
    → Claude reads the PDF
    → Clean Markdown, Excalidraw diagrams, or whatever you need

No API keys. No cloud access. Everything runs locally. The desktop app handles syncing; this server just reads the files it produces.

Related MCP server: obsidian-ai-curator

Prerequisites

• reMarkable desktop app — installed and synced (download)

• macOS — the server reads from the standard macOS cache path (Linux support is possible but untested)

• Python 3.12+

• uv — Python package manager (install)

• cairo — system graphics library for SVG→PDF rendering:

  brew install cairo

Installation

git clone https://github.com/sambt94/remarkable-mcp.git
cd remarkable-mcp
uv sync

MCP Registration

Add the server to your Claude Code MCP configuration. See mcp.example.json for the full template, or add this to your .mcp.json:

{
  "mcpServers": {
    "remarkable": {
      "type": "stdio",
      "command": "/bin/bash",
      "args": [
        "-c",
        "export DYLD_LIBRARY_PATH=/opt/homebrew/lib; exec uv --directory '/path/to/remarkable-mcp' run server.py"
      ]
    }
  }
}

Replace /path/to/remarkable-mcp with the actual path to your cloned repo.

The DYLD_LIBRARY_PATH is needed so Python can find the Homebrew-installed cairo library on macOS.

Tools

The server exposes 6 tools:

Page selection

remarkable_render_pages supports flexible page selection:

# Last 5 pages of a document
remarkable_render_pages(doc_id="<uuid>", last_n=5)

# First 3 pages
remarkable_render_pages(doc_id="<uuid>", first_n=3)

# Specific pages (0-indexed)
remarkable_render_pages(doc_id="<uuid>", page_indices=[0, 2, 4])

# All pages (no selection args)
remarkable_render_pages(doc_id="<uuid>")

Priority: page_indices > last_n > first_n > all pages.

Usage with Claude

Once registered, Claude can access your reMarkable notebooks directly:

"Transcribe the last 3 pages of my journal"

"Find my notebook called 'Architecture Notes' and render page 5"

"What documents do I have on my reMarkable?"

The rendered PDFs are saved to /tmp/remarkable-renders/ and can be cleaned up with remarkable_cleanup_renders.

Companion Skills

The skills/ directory contains Claude Code skill definitions that wrap the MCP tools into complete workflows:

• remarkable-transcribe.md — Transcribe handwritten notes to clean Markdown

• remarkable-diagram.md — Convert hand-drawn diagrams to interactive Excalidraw files

To use these, copy the skill files into your ~/.claude/skills/ directory (or symlink them).

Architecture

remarkable-mcp/
├── server.py                 # MCP server entry point (FastMCP, stdio transport)
├── remarkable_client.py      # Client library (cache reading, rendering pipeline)
├── skills/
│   ├── remarkable-transcribe.md   # Handwriting → Markdown skill
│   └── remarkable-diagram.md      # Diagram → Excalidraw skill
└── tests/
    ├── conftest.py           # Synthetic cache fixtures
    ├── test_remarkable_client.py  # Unit tests
    ├── test_server.py        # Integration tests
    └── test_e2e.py           # End-to-end stdio tests

The rendering pipeline:

1. rmc parses reMarkable's proprietary .rm binary format (v6) into SVG

2. cairosvg converts SVG to PDF

3. pypdf merges per-page PDFs into a single document

4. Claude reads the PDF and does whatever you need — transcription, diagram interpretation, summarization

Tests

# All tests
uv run pytest tests/ -v

# By category
uv run pytest tests/ -m unit          # unit tests (synthetic cache)
uv run pytest tests/ -m integration   # tool registration and response shapes
uv run pytest tests/ -m e2e           # full stdio transport

Tests use synthetic fixtures — no real reMarkable device or cache required.

License

MIT — see LICENSE.