Dart项目管理
STDIO提供任务管理和文档处理的Dart服务器
提供任务管理和文档处理的Dart服务器
A Model Context Protocol (MCP) server implementation for Dart, providing task management, document handling, and workspace organization capabilities through MCP tools.
pip install dart-sdk
)To install Dart MCP Server for Claude Desktop automatically via Smithery:
npx -y @smithery/cli install @jmanhype/dart-mcp-server --client claude
git clone https://github.com/jmanhype/dart-mcp-server.git cd dart-mcp-server
npm install
# Create and activate virtual environment python -m venv .venv source .venv/bin/activate # On Windows: .venv\Scripts\activate # Install Dart SDK pip install dart-sdk
# Copy example environment file cp .env.example .env # Edit .env with your configuration # Required: DART_TOKEN # Optional: PYTHONPATH (path to dart sdk)
npm run build
npm start
# Watch for TypeScript changes npm run dev # Run tests npm test
Create a .env
file with the following variables:
# Required: Your Dart API token DART_TOKEN=your_dart_token_here # Optional: Path to your Dart SDK installation PYTHONPATH=/path/to/dart/sdk # Optional: Python executable path (defaults to system Python) PYTHON_PATH=/path/to/python
create_task
: Create new tasks with title, description, priority, etc.update_task
: Update existing tasks' status, title, descriptionget_default_status
: Get default status DUIDsget_default_space
: Get default space DUIDget_dartboards
: List available dartboardsget_folders
: List folders in a spacecreate_folder
: Create new folderscreate_doc
: Create new documents or reportscreate_space
: Create new workspacesdelete_space
: Delete existing workspacesIf you encounter issues:
Verify Python environment:
python --version pip list | grep dart
Check Dart SDK installation:
python -c "import dart; print(dart.__version__)"
Verify environment variables:
echo $DART_TOKEN echo $PYTHONPATH
MIT License
PyPI Supported Python Versions License
Dart is Project Management powered by AI.
dart-tools
is the Dart CLI and Python Library. It enables direct integration with Dart through a terminal CLI or through Python.
In the terminal, install by running
pip install dart-tools
Start off by setting up authentication with
dart login
Then, you can create a new task with a command along the lines of
dart createtask "Update the landing page" -p0 --tag marketing
which will make a new task called 'Update the landing page' with priority 'Critical' (i.e. P0) and with the 'marketing' tag.
You can explore all of these options and many more with dart --help
or the more specific help for subcommands, in this case dart createtask --help
.
Another common workflow is to updating a preexisting task. To do this, run something like
dart updatetask [DUID] -s Done
This command will mark the referenced task 'Done'. Here [DUID]
is meant to be replaced (including the brackets) with the 'Dart ID' of an existing task. You can get a DUID from any existing task in a number of ways, such as by copying it from the end of a task's URL or by clicking the '...' button in a task page in Dart and then choosing 'Copy ID'.
First, set up authentication. Run dart login
in the terminal for an interactive process, or visit your Dart profile and then run dart.login(token)
or save the token into the DART_TOKEN
environment variable.
Then, you can run something like
import os from dart import create_task, is_logged_in, update_task # Check that auth is set up and stop if not, can remove this once everything is set up is_logged_in(should_raise=True) # Create a new task called 'Update the landing page' with priority 'Critical' (i.e. p0) and with the 'marketing' tag new_task = create_task( "Update the landing page", priority_int=0, tag_titles=["marketing"] ) # Update the task to be 'Done' update_task(new_task.duid, status_title="Done")
The Model Context Protocol (MCP) server implementation enables AI assistants (like Claude) to interact with Dart through standardized tools. This allows for seamless integration of AI capabilities with Dart's task management system.
# Clone the repository git clone https://github.com/its-dart/dart-tools.git cd dart-tools/dart/mcp # Install dependencies npm install # Set up Python environment python -m venv .venv source .venv/bin/activate # On Windows: .venv\Scripts\activate pip install dart-tools # Configure environment cp .env.example .env # Edit .env with your DART_TOKEN
The server provides these MCP tools:
For detailed documentation, see MCP Server README.
Almost anything that can be done in Dart can be done with the Python library, but there are not convenient wrapper functions for everything. For most advanced usage, the best thing to do is to get in touch with us and we can help.
However, if you want to explore on your own, the client is well-typed, so you can simply explore the code to see what is possible. All updates will go through the the dart.transact
function.
As an example, you could run something akin to update_task
with
from dart import ( Dart, Operation, OperationKind, OperationModelKind, TaskUpdate, TransactionKind, ) # Initialize the inner client dart = Dart() # Prepare the update operation task_update = TaskUpdate( duid="[DUID]", size=5, ) task_update_op = Operation( model=OperationModelKind.TASK, kind=OperationKind.UPDATE, data=task_update, ) # Call the operation transactionally to perform the update response = dart.transact([task_update_op], TransactionKind.TASK_UPDATE)
Contributions are welcome! Please open an issue or submit a pull request.
This project is licensed under the MIT License.