celery-monitor

What problem does it solve?

Celery systems often fail silently or slow down due to backlogs, dead workers, or stuck/failed tasks, and it can be difficult to pinpoint the root cause quickly.

Core Features & Use Cases

• Queue depth inspection: Detect pending task backlog by checking the Redis list backing the default queue.
• Worker liveness checks: Confirm whether workers are alive by looking for heartbeat key patterns.
• Failure and stuck-task triage: Identify failed tasks via task-meta states and surface long-running STARTED tasks that may be stuck, then correlate with container logs when needed.
• Result backend hygiene: Estimate whether task result entries are accumulating by counting task-meta keys and understanding expiry behavior.

Quick Start

Use the celery-monitor skill to inspect queue depth, worker heartbeats, and failed or stuck tasks in Redis MCP for your Celery deployment.