Celery Flower MCP

- Monitorworkers, tasks, and queues in real time
- Controlworker pools — grow, shrink, autoscale, restart, shut down
- Manage tasks— apply, revoke, abort, set timeouts and rate limits
- Inspect queues— check depths, add/remove consumers

All 21 Flower API endpoints are covered.

FLOWER_URL=http://localhost:5555 uvx celery-flower-mcp

git clone https://github.com/Darius1223/celery-flower-mcp cd celery-flower-mcp uv sync uv run python -m source.main

Add to~/Library/Application Support/Claude/claude_desktop_config.json:

make fmt # auto-format with ruff make lint # lint with ruff make typecheck # type-check with mypy (strict) make test # run 49 unit tests make cov # unit tests + coverage report make all # fmt + lint + typecheck

The test suite is split into two layers:

Unit tests(tests/) — fast, no external dependencies, usepytest-httpxto mock HTTP calls:

make test # or uv run pytest tests/ -m "not integration"

Integration tests(tests/integration/) — run against a real Flower instance backed by Redis and a live Celery worker, all managed by Docker Compose:
- Builds and starts the Docker Compose stack (docker-compose.test.yml) — Redis → Celery worker → Flower
- Waits for Flower's/healthcheckendpoint to return OK
- Runs the 16 integration tests againsthttp://localhost:5555
- Tears down the stack when done

The stack is defined indocker-compose.test.yml. The worker and Flower images are built fromtests/integration/Dockerfile.workerandtests/integration/Dockerfile.flower.

To start the stack manually for exploratory testing:

docker compose -f docker-compose.test.yml up -d --build # run tests, explore, etc. make integration-down # stop + remove volumes

Integration tests usepytest.mark.asyncio(loop_scope="session")so all tests share one event loop — this avoidsRuntimeError: Event loop is closedwhen httpx transports are cleaned up across test boundaries on Python 3.14.

This is a web browser that enables your coding agent, such as Claude Code, to visit websites on your behalf and assist you in identifying bugs or creating UI test cases.

Enable AI Agents to fix build failures from CircleCI.

Access and interact with Harness platform data, including pipelines, repositories, logs, and artifact registries.

Scout's official MCP pipes error, trace and metric data from production to your AI agent

A server that provides background process management capabilities, enabling LLMs to start, stop, and monitor long-running command-line processes.

Stateful health monitoring, diagnostics, and web attestation for AI agents. 11 MCP tools. Free Founder's Beta

Breaking Change Alerts for Humans and AI Agents.

Paid remote MCP for CLI tool MCP, structured receipts, usage logs, and audit-ready evidence for agent and CI workflows.

Interact with the CODING DevOps platform for managing projects and work items.

Remote MCP server that gives your AI agent a dead-man's-switch: it calls create_heartbeat, pings on schedule, and emails you the instant an agent or cron loop stops checking in.