Supatask

Supatask is a Redis-based local task manager that exposes an MCP server interface for AI assistants to create, read, update, and delete tasks, as well as query logs. It enables seamless automation and high‑level task orchestration through a standardized MCP HTTP transport, letting you integrate planning and execution flows directly with your own AI or automation tools.

How to use

You interact with Supatask through an MCP client that speaks JSON‑RPC over HTTP to the /mcp/ endpoint. Start by initializing the MCP session, then discover available tools and call them to manage tasks or read logs. Use the tools to create new tasks with tags, read task details, list tasks with filters, update tasks, delete tasks, and retrieve activity or system logs. You can start and stop time tracking for tasks and have access to a time history for each task.

Key workflows you can implement with an MCP client include creating a task with specific tags, listing tasks filtered by status or tags, updating a task’s title, description, or tags, starting and stopping a task timer, and fetching recent logs for auditing or debugging. The MCP interface provides a clean abstraction so your AI assistant can orchestrate work without needing to understand the underlying HTTP endpoints.

How to install

Prerequisites you need on your machine before running Supatask are Docker and Docker Compose. Ensure you have a modern terminal and network access to pull images from the Docker registry.

Step 1: Start the application locally with Docker Compose. This will bring up the backend service, Redis, and any frontend or auxiliary services defined for the stack.

Step 2: Verify that the services are running and healthy. You should be able to reach the MCP endpoint and the web UI after startup.

Step 3: Optional – configure environment variables for fine‑grained behavior. The system reads REDIS_URL and LOG_LEVEL to connect to Redis and control log verbosity.

Configuration and usage notes

The MCP server listens for JSON‑RPC 2.0 requests at /mcp/ and supports an HTTP Streamable transport. It exposes the following tools for AI integration: create_task, read_task, list_tasks, update_task, delete_task, and get_logs. Use these tools to perform common task management operations and to retrieve logs for auditing.

Environment variables you may configure include REDIS_URL for Redis connectivity and LOG_LEVEL for log verbosity. If you change these, restart the services so the changes take effect.

Troubleshooting tips

If the MCP endpoint is not reachable, confirm Docker is running, the containers are up, and the /mcp/ URL is accessible on the configured host and port. Check Redis connectivity via the REDIS_URL setting and inspect log output to identify authentication, connection, or timeout issues.

Available tools

create_task

Create tasks with tags and initial attributes such as title, description, status, and tags.

read_task

Read detailed information for a specific task by ID, including time tracking history.

list_tasks

List tasks with optional filters like status, tags, created_after, and created_before.

update_task

Update task fields such as title, description, status, and tags.

delete_task

Delete a task by its ID.

get_logs

Retrieve activity and system logs for auditing and debugging.