MCP Server MCP

Incident Response MCP for Apache Airflow

By github 👁 1 views ▲ 0 votes

Diagnose Airflow failures from UI links with bounded logs and optional recovery actions.

Homepage Verified source GitHub

Install

uvx apache-airflow-mcp-server

Configuration Example

{
  "remotes": [],
  "packages": [
    {
      "registryType": "pypi",
      "registryBaseUrl": "https://pypi.org",
      "identifier": "apache-airflow-mcp-server",
      "version": "1.0.2",
      "runtimeHint": "uvx",
      "transport": {
        "type": "stdio"
      },
      "environmentVariables": [
        {
          "description": "Base URL of the Airflow instance.",
          "isRequired": true,
          "placeholder": "https://airflow.example.com",
          "name": "AIRFLOW_MCP_HOST"
        },
        {
          "description": "Bearer token. Provide this or both username and password; it takes precedence when set.",
          "isSecret": true,
          "name": "AIRFLOW_MCP_TOKEN"
        },
        {
          "description": "Basic-auth username. Required with password when token is omitted.",
          "name": "AIRFLOW_MCP_USERNAME"
        },
        {
          "description": "Basic-auth password. Required with username when token is omitted.",
          "isSecret": true,
          "name": "AIRFLOW_MCP_PASSWORD"
        },
        {
          "description": "The directory install uses the Airflow 3 client. See the README for Airflow 2 setup.",
          "value": "v2",
          "name": "AIRFLOW_MCP_API_VERSION"
        },
        {
          "description": "Expose only read tools unless write access is deliberately selected.",
          "format": "boolean",
          "default": "true",
          "choices": [
            "true",
            "false"
          ],
          "name": "AIRFLOW_MCP_READ_ONLY"
        }
      ]
    }
  ]
}

mcp model-context-protocol pypi