ZenML MCP server

This is a small helper that connects your AI assistant to ZenML, the platform teams use to build and run machine learning pipelines. Once it is set up, you can ask your AI plain questions about your pipelines, runs, models, and deployments instead of clicking around the ZenML dashboard. It is handy for data scientists, ML engineers, and anyone who works with ZenML pipelines and wants quick answers or a quick way to kick off a run.

What this MCP server does

You ask your AI a question about ZenML, like which pipeline runs failed today. Your AI passes that request to this helper program. The helper talks to your ZenML server using the credentials you set up. ZenML sends back the information, and the helper hands it to your AI, which explains it to you in the chat. It can also start a new pipeline run for you when you ask it to.

Click to zoom

What you can do with it

• List your pipelines, pipeline runs, and their statuses
• Read the logs and code of individual pipeline steps
• Check which deployments are running and see their logs
• Look up models, model versions, and artifacts
• Browse stacks, stack components, and service connectors
• Trigger a new pipeline run from a snapshot
• See your projects, tags, and builds

Try asking your AI

• “Show me the pipeline runs that failed in the last 24 hours”
• “What is the status of my latest deployment?”
• “Get the logs for the training step of my most recent run”
• “Trigger a run of my nightly training snapshot”

What it gives back to you

You get answers in plain chat text: lists of runs with their statuses, short summaries of deployments, or log lines from a step. When you ask it to trigger a run, it tells you whether the request was accepted and gives you a run reference. If a log is very long, it shows the newest lines and tells you that older ones were left out.

Before you start

Who it's for

Data scientists, ML engineers, and MLOps teams who already use ZenML and want to ask questions or start runs without leaving their AI chat.