Apache Airflow

Integrates with Apache Airflow clusters through REST API to provide complete workflow management including DAG operations, task monitoring, pool and variable management, XCom data access, and performance analytics with event logging and import error tracking.

MCP server implementation by call518 that provides AI assistants with complete access to Apache Airflow cluster management and monitoring through REST API integration, built using Python with FastMCP and aiohttp for high-performance async operations. The implementation offers 43 tools covering DAG management (listing, triggering, pausing), cluster health monitoring, pool and variable management, task instance tracking with comprehensive filtering, XCom data access, configuration management, and detailed analytics including event logs, import errors, and performance metrics. Built with modern async HTTP architecture featuring connection pooling, persistent sessions, and optimized pagination defaults, it serves DevOps teams managing Airflow workflows, data engineers monitoring pipeline health, and organizations requiring conversational access to Airflow operations without direct UI interaction, with support for both stdio and HTTP transport methods plus Docker deployment options.