dart
Dart AI Model Context Protocol (MCP) server
claude mcp add --transport stdio its-dart-dart-mcp-server npx -y dart-mcp-server@latest \ --env DART_TOKEN="dsa_..."
How to use
The Dart MCP Server exposes a local Model Context Protocol (MCP) endpoint for Dart, enabling AI assistants to manage tasks and documents within a Dart workspace. It provides a suite of prompts and tools designed to interact with tasks and documents in a structured, JSON-based flow. You can run the server using npx, then connect your Dart-enabled clients to fetch configuration data, list or create tasks, update documents, and perform common workspace operations via the MCP tools. The server's tooling includes task management (get_config, list_tasks, create_task, get_task, update_task, delete_task, add_task_comment) and document management (list_docs, create_doc, get_doc, update_doc, delete_doc), all returning structured JSON responses for easy consumption by clients.
To use the server, configure your MCP client (Claude, Cursor, Windsurf, Cline, Windsurf, or any other client) with the following setup, replacing the DART_TOKEN with your actual Dart profile token. Once configured, you can start issuing prompts and tool calls to manage tasks and docs through the Dart MCP Server. The npx-based setup is the recommended starting point, with an optional Docker variant if you prefer containerized execution.
How to install
Prerequisites:
- Node.js and npm (npm comes with Node.js)
- Access to your Dart token from your Dart profile
- Optional: Docker if you prefer the containerized variant
Install and run the MCP server locally using npx:
- Ensure Node.js and npm are installed. You can download from https://nodejs.org
- Acquire your Dart token from your Dart profile (DART_TOKEN)
- Start the MCP server via npx:
npx -y dart-mcp-server@latest
Configure your MCP client (example for the standard Dart setup):
- Create or edit your MCP config to include the Dart server with the token:
{
"mcpServers": {
"Dart": {
"command": "npx",
"args": ["-y", "dart-mcp-server@latest"],
"env": {
"DART_TOKEN": "dsa_..."
}
}
}
}
If you prefer Docker:
- Build the container image (provided in the setup):
# Build the image (already handled in the guide, ensure you have Docker installed)
docker build -t mcp/dart .
- Run with the token exposed as an environment variable:
docker run -e DART_TOKEN -i mcp/dart
- Point your MCP client to the running container using the same mcpServers configuration as above, but with command set to bash and appropriate docker run shell command as shown in the guide.
Additional notes
Tips and notes:
- The server requires your DART_TOKEN to authorize requests to the Dart API; keep this token secure and do not commit it in public repos.
- The official README indicates a deprecated local server in favor of the hosted Dart MCP server, but the npx option remains available for local testing.
- If you encounter port or network issues when running locally, ensure your firewall allows the script to establish outgoing connections as required by the Dart API.
- Docker-based setup is provided as an alternative; use the docker variant if you prefer containerized execution or wish to isolate dependencies.
- The npm package name for this server is dart-mcp-server; you can install or reference it directly via npm or npx.
Related MCP Servers
outline
It's an MCP server... for Outline (the documentation platform!)
CanvasMCPClient
Canvas MCP Client is an open-source, self-hostable dashboard application built around an infinite, zoomable, and pannable canvas. It provides a unified interface for interacting with multiple MCP (Model Context Protocol) servers through a flexible, widget-based system.
docmole
Dig through any documentation with AI - MCP server for Claude, Cursor, and other AI assistants
guildbridge
🏰 Remotely hosted MCP server for Discord
obsidian
MCP server for Obsidian vault management - enables Claude and other AI assistants to read, write, search, and organize your notes
xgmem
Global Memory MCP server, that manage all projects data.