reMarkable MCP Server · wavyrai — Claude Code MCP server: read and search notes
Gives an AI agent access to a reMarkable library to browse documents, search their contents and extract handwriting with OCR.
claude mcp add rm-mcp -e REMARKABLE_TOKEN=<REMARKABLE_TOKEN> -- uvx rm-mcpWhat reMarkable MCP Server does
The server lets an agent browse folders, read documents, search across files and retrieve page images. It extracts text from PDFs, EPUBs and typed notes. Reading a document indexes its text for later full-text searches.
Handwriting OCR uses sampling: page images go to the MCP client’s AI model for text extraction. The client must support sampling, while OCR needs no separate API key. The server connects through the reMarkable Cloud API, so a reMarkable Connect subscription and token are required. Read and search tools are read-only; folder management allows renaming, moving and creating folders, but not deleting files.
What it helps with
- search a reMarkable library for a specific passage
- extract text from PDFs and EPUBs
- turn handwritten notes into searchable text
- read selected pages from a document
- view recent documents and page images
Example prompts for the agent
- Search my reMarkable library for notes about the project deadline.
- Read pages 1–3 of
Research Paperand extract their text. - Use OCR to transcribe the handwritten entries in
Journal.
How to install reMarkable MCP Server
Platform CLI
claude mcp add rm-mcp -e REMARKABLE_TOKEN=<REMARKABLE_TOKEN> -- uvx rm-mcpReplace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from rm-mcp <command>
Platform CLI
codex mcp add wavyrai-remarkable-mcp-server --env REMARKABLE_TOKEN=<REMARKABLE_TOKEN> -- uvx rm-mcpReplace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from rm-mcp <command>
Config file
{
"mcpServers": {
"wavyrai-remarkable-mcp-server": {
"command": "uvx",
"args": [
"rm-mcp"
],
"env": {
"REMARKABLE_TOKEN": "<REMARKABLE_TOKEN>"
}
}
}
}Replace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from rm-mcp <command>
Config file
{
"servers": {
"wavyrai-remarkable-mcp-server": {
"type": "stdio",
"command": "uvx",
"args": [
"rm-mcp"
],
"env": {
"REMARKABLE_TOKEN": "<REMARKABLE_TOKEN>"
}
}
}
}Replace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from rm-mcp <command>
Platform CLI
gemini mcp add -e REMARKABLE_TOKEN=<REMARKABLE_TOKEN> wavyrai-remarkable-mcp-server uvx rm-mcpReplace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from rm-mcp <command>
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about reMarkable MCP Server
How do I install reMarkable MCP Server in Claude Code?
Run claude mcp add rm-mcp -e REMARKABLE_TOKEN=<REMARKABLE_TOKEN> -- uvx rm-mcp. You need a reMarkable token to connect.
What does reMarkable MCP Server do?
It gives an AI agent tools to browse a reMarkable library, read documents, search their contents and retrieve page images. Sampling OCR can extract handwriting.
What does handwriting OCR require?
The MCP client must support sampling, and the server needs a reMarkable token. OCR uses the client’s AI model, so it does not require a separate API key.
Similar MCP servers
Everything for Notes, tasks and projects →reMarkable MCP Server by other authors comes first.
