Readwise-Reader

This MCP server provides an integration with Readwise Reader accessible via Claude and other MCP clients. It exposes a set of tools to save, list, update, and delete documents from Readwise Reader, along with tag management and advanced filtering options. The server internally uses the Readwise Reader API, handles authentication via a token provided in the environment variable READWISE_TOKEN, and converts content into an LLMin/lib-friendly format to support downstream processing. Typical usage involves invoking tools to save documents (by URL or HTML), retrieve and filter documents with metadata, and modify existing entries as needed. The content is optimized for LLM workflows, with HTML content converted to clean text where appropriate.

Prerequisites:

• Node.js and npm installed on your system
• Access to Readwise Reader and a Readwise API token

Steps:

1. Install dependencies and build the MCP server: npm install npm run build

2. Obtain a Readwise access token from Readwise: https://readwise.io/access_token

3. Configure the MCP server in Claude Desktop (or your MCP client) by pointing to the built index.js and setting the token: { "mcpServers": { "readwise-reader": { "command": "node", "args": ["/path/to/your/reader_readwise_mcp/dist/index.js"], "env": { "READWISE_TOKEN": "your_readwise_access_token_here" } } } }

4. Replace "/path/to/your/reader_readwise_mcp" with the actual path to this project directory and restart Claude Desktop or your MCP client.

Environment variable: READWISE_TOKEN must be set to a valid Readwise access token. Token is stored in the MCP configuration and used for authenticating all API requests to Readwise Reader. Be mindful of rate limits (default 20 requests/minute; document create/update up to 50 requests/minute) and use pagination when listing large document sets. When listing documents, you can enable withHtmlContent and withFullContent with caution due to performance considerations. This server returns complete document information including content, metadata, and timestamps, suitable for downstream analysis and chaining with LLM workflows.