MCP server · Developer tools
Airflow MCP server
by us-all
Lets your AI check Apache Airflow DAGs, runs, and task logs, and trigger or clear runs when you allow it.
This is a small helper that connects your AI assistant to Apache Airflow, the tool many data teams use to schedule and run their data jobs. Once it is set up, you can ask your AI things like why a job failed or how healthy your pipelines are, instead of clicking around the Airflow website. It is handy for data engineers, analysts, and anyone who keeps an eye on scheduled jobs.
What is an MCP server? The 30-second version
On its own, your AI can only chat with you. An MCP server is a small helper program that gives your AI a new skill or a connection to an app, and here it connects your AI to your Airflow installation. When you ask a question, your AI sends it to this helper, the helper talks to Airflow for you, and the answer comes back in the chat. You do not need to understand how it works inside; you just need it set up once.
What this MCP server does
You ask your AI a question about your Airflow jobs, like which runs failed today. The AI passes that request to this helper, which logs in to your Airflow server and asks it for the information. Airflow sends back the list of DAGs, runs, task details, or log text, and the helper hands it to your AI. Your AI then writes it up for you in plain language. If you turn on write mode, the helper can also ask Airflow to start a new run or clear a task so it runs again.
Click to zoomWhat you can do with it
- List your active DAGs, with filters by tag or name
- Show recent runs of a DAG and their states
- List the task instances inside a specific run
- Fetch the tail of a task log so you can see the error
- Roll up DAG health with success rate, average duration, and the last failure
- Trigger a new DAG run (only if write mode is on)
- Clear a task instance so it runs again (only if write mode is on)
Try asking your AI
- “Which DAGs failed in the last 24 hours?”
- “Show me the recent runs of the daily_sales_load DAG and their states”
- “Why did the orders_etl DAG fail last night? Pull the failing task log.”
- “Give me a health summary for all my DAGs this week”
What it gives back to you
You get back plain answers in the chat: lists of DAG names, run states and timestamps, task names with their status, and short excerpts of log text. The health rollup comes back as numbers, like a success rate and an average duration, plus the name of the last failed run. If you use a write tool, it tells you whether the trigger or clear was accepted. Everything shows up as normal chat text you can read or copy.
Before you start
What you need
- An Apache Airflow 3.x server you can reach (Airflow 2.x is not supported by this version)
- The web address of that Airflow server
- An Airflow username and password that can log in
- Node.js installed on your computer, plus the Claude desktop app or another MCP client
Good to know
By default it is read-only, but if you turn on write mode the AI can start new DAG runs and clear tasks, which changes what your pipelines do, so switch that on only when you mean to.
Install it with your AI
Add Airflow MCP server to your AI, no technical skills needed
You don't install anything by hand. You copy one prompt, paste it into an AI that can work on your computer, and it checks, installs and connects the server for you, asking you when it needs something.
Sign in to get the install prompt
Members get a ready-made prompt that lets the Claude desktop app check Airflow MCP server, install it and connect it for them, step by step. You don't need any technical skills: you copy, paste and answer a few questions. Your connected AI can also find and install any of the 4,066 MCP servers here for you.
Who it's for
Data engineers, analytics engineers, and data platform teams who run scheduled jobs in Airflow and want quick answers without opening the Airflow web UI.
