Raindrop MCP Server

An MCP (Model Context Protocol) server that exposes Raindrop.io's bookmark management API as tools for AI assistants.

Features

• Collections: List, create, update, and delete collections with full path support
• Bookmarks: Search, create, update, move, and delete bookmarks
• Tags: List, rename, delete, merge, and add/remove tags on bookmarks
• Stats: Get filter statistics for collections

Prerequisites

• Python 3.12+
• A Raindrop.io account with a Personal Access Token

Getting Your Raindrop API Token

1. Log in to Raindrop.io
2. Go to Developer Integrations
3. Select Create new app and enter an app name
4. Click the name of your app and create a test token
5. Copy the test token — you'll need it for the .env file

Setup

1. Clone and configure

git clone <repository-url>
cd raindrop_mcp_server

2. Create your environment file

cp .env.example .env

Edit .env with your values:

Variable	Description	Default
RAINDROP_ACCESS_TOKEN	Required. Access Token	—
HOST	Host to bind the MCP server	0.0.0.0
PORT	Port to bind the MCP server	8000
COLLECTION_CACHE_TTL_SECONDS	Cache TTL	86400
LOG_LEVEL	Logging level	INFO
RAINDROP_API_BASE	Raindrop API base URL	https://api.raindrop.io/rest/v1

3. Install dependencies with uv

uv sync

To install test dependencies as well:

uv sync --all-extras

Running the Server

uv run python app.py

The server starts an MCP endpoint on the configured host and port using the streamable-http transport.

Testing the Server

After starting the server, verify it works with the included test script:

bash test_list_collections.sh

The script reads your .env file, initializes an MCP session, and calls list_collections to confirm the server is responding correctly.

Running Tests

The tests are unit tests that do not contact the live Raindrop API. They require the test extra to be installed:

uv run --all-extras pytest -q

Or with pytest directly after installing:

uv sync --all-extras
uv run pytest -q

Docker

docker build -t raindrop-mcp .
docker run --env-file .env -p 8000:8000 raindrop-mcp

Available Tools

Tool	Description
list_collections	Return all Raindrop collections using full paths
refresh_collections_cache	Force a refresh of the cached collection tree
create_collection	Create a collection at a full path
update_collection	Update a collection (rename, move, change settings)
delete_collection	Delete a collection by full path
list_bookmarks	Return all bookmarks in a collection
search_bookmarks	Search bookmarks by metadata
get_bookmark	Get a single bookmark by ID or URL
create_bookmark	Create a new bookmark in a collection
update_bookmark	Update bookmark metadata
move_bookmark	Move a bookmark to another collection
delete_bookmark	Delete a bookmark by ID or URL
list_tags	Return tag vocabulary, optionally scoped to a collection
rename_tag	Rename a tag globally or within a collection
delete_tag	Delete a tag globally or within a collection
merge_tags	Merge multiple tags into one
add_tags	Add tags to a bookmark without removing existing ones
remove_tags	Remove tags from a bookmark
collection_stats	Return filter counts for a collection or all collections

License

MIT License. See LICENSE for details.