raindrop-mcp

What is raindrop-mcp?

An MCP (Model Context Protocol) server that integrates with Raindrop.io, allowing AI assistants to search, create, and manage bookmarks via the Raindrop.io API. It is intended for developers and users who want to give LLM tools access to their Raindrop.io bookmarks and collections.

How to use raindrop-mcp?

1. Create an access token on Raindrop.io by creating an application and generating a test token. 2. Configure the server as a stdio MCP server by setting the RAINDROP_ACCESS_TOKEN environment variable. 3. Run via npx -y raindrop-mcp. Alternatively, start it as an SSE server with --sse or a streamable HTTP server with --http.

Key features of raindrop-mcp

• Search bookmarks by query
• Create new bookmarks
• Retrieve all collections
• Supports stdio, SSE, and streamable HTTP transports

Use cases of raindrop-mcp

• Asking an AI to find a saved article or link
• Adding a new bookmark from within a chat interface
• Listing all collections to organize bookmarks
• Integrating bookmark management into AI-powered workflows

FAQ from raindrop-mcp

What is required to use raindrop-mcp?

You need a Raindrop.io account and a generated access token from the Raindrop.io developer settings. The server runs via npx so Node.js must be installed.

What transports does raindrop-mcp support?

It supports stdio (default), SSE (--sse), and streamable HTTP (--http).

Where are the bookmarks stored?

All data lives in your Raindrop.io account. The server only interacts with Raindrop.io’s API and does not store data locally.

How do I provide authentication?

Set the environment variable RAINDROP_ACCESS_TOKEN to your Raindrop.io test token. This is the only authentication mechanism described.

Are there any known limitations?

The README does not list any limitations. The server currently offers three capabilities: search, create bookmarks, and list collections.