celery-flower-mcp is a Model Context Protocol (MCP) server that transforms the comprehensive Celery Flower REST API into AI-actionable tools. By pointing this server at an existing Flower instance, AI assistants like Claude, Cursor, or Windsurf gain the ability to monitor workers, tasks, and queues in real-time, control worker pools (grow, shrink, autoscale, restart, shut down), manage individual tasks (apply, revoke, abort, set timeouts and rate limits), and inspect/manage queues (check depths, add/remove consumers). It provides full coverage of all 21 Flower API endpoints, offering a powerful interface for AI-driven automation and oversight of Celery operations.
Key Features
01Full API coverage exposing every Celery Flower REST endpoint as an MCP tool
02Typed configuration using Pydantic Settings with .env file support
03Asynchronous architecture built on httpx and FastMCP
04Robust test suite including unit and integration tests against a real Flower instance
05Strict typing with mypy in strict mode for high code quality
063 GitHub stars
Use Cases
01Automating the monitoring and management of Celery workers and their pools (scaling, restarting).
02Enabling AI assistants to apply, revoke, and manage individual Celery tasks programmatically.
03Allowing AI-driven inspection and modification of Celery queue depths and consumer configurations.