This MCP server implements a ClickUp integration following the MCP protocol. It exposes tools that interact with ClickUp's API and allows clients to enumerate available tools and invoke specific capabilities, such as listing authorized teams or retrieving data from ClickUp resources. Clients can connect using either SSE or streaming HTTP transports, then perform session initialization and call tools by name to obtain structured responses. To get started, ensure you have a valid ClickUp API token and start the server with the appropriate environment variable or a .env file containing CLICKUP_API_TOKEN. Once the server is running, you can connect with an MCP client to discover the available tools and invoke them to interact with ClickUp resources (e.g., Teams, Spaces).

The included examples show how to connect via SSE and streaming HTTP transports, initialize a session, list tools, and call a tool like get_authorized_teams. The client code demonstrates how to read tool metadata and receive results through the MCP protocol, enabling AI applications to orchestrate ClickUp operations through standardized MCP calls.

Prerequisites:

• Python 3.13 or higher
• pip (Python package installer)

1. Create a virtual environment (recommended): python -m venv venv source venv/bin/activate # macOS/Linux .\venv\Scripts\activate # Windows

2. Install the ClickUp MCP Server package: pip install clickup-mcp-server

3. (Optional) Copy or configure environment variables in a .env file:

   .env

   CLICKUP_API_TOKEN=your_clickup_token_here

4. Run the server: clickup-mcp-server --token YOUR_CLICKUP_API_TOKEN

Note: The CLI entry point shown (clickup-mcp-server) uses the environment token for authentication. If you provide the token via a .env file, you can start with --env /path/to/.env to load it.

Tips and common issues:

• Ensure your ClickUp API token has the required scopes for the resources you intend to access (Teams, Spaces, etc.).
• If you store configuration in a .env file, use the --env option to point to that file when starting the server.
• The MCP server supports multiple transport methods (SSE and streaming HTTP). Use the client implementations demonstrated in the README to connect accordingly.
• If you encounter connection errors, verify that the server port (default 8000) is not blocked by a firewall and that the host binding is accessible by your client.
• You can override the default port/host as needed (e.g., --host 0.0.0.0 --port 8001).
• For debugging, run with a verbose logger if available in the CLI or client to capture request/response details.