mcp-zenmlMCP server to connect an MCP client (Cursor, Claude Desktop etc) with your ZenML MLOps and LLMOps pipelines

— (0)

Reviewed by Daniel Nikulshyn·Updated July 2026

Overview

The mcp-zenml project implements a Model Context Protocol (MCP) server for interacting with the ZenML API. It provides a standardized way to connect AI models to different data sources and tools, acting as a bridge between MCP clients (like Cursor, Claude Desktop) and ZenML MLOps and LLMOps pipelines. The server exposes core read functionality from the ZenML server, allowing access to live information about various entities, such as users, stacks, pipeline execution, and artifacts. It also enables triggering new pipeline runs using snapshots or run templates.

Key features

MCP server for interacting with ZenML API
Access to core read functionality from ZenML server
Triggering new pipeline runs using snapshots or run templates
Exposure of tools for pipeline execution, deployment, and serving

Pricing

Model: Free
Category: MCP Servers
Rating: No reviews yet

Use cases

Connecting MCP clients to ZenML pipelines

Use the mcp-zenml server to connect MCP clients like Claude Desktop to ZenML MLOps and LLMOps pipelines, enabling standardized access to data sources and tools.

Triggering pipeline runs

Utilize the mcp-zenml server to trigger new pipeline runs using snapshots or run templates, streamlining the execution of machine learning workflows.

Pros & Cons

Pros

Standardized protocol for connecting AI models to data sources and tools
Provides live information about core ZenML entities
Enables triggering new pipeline runs using snapshots or run templates

Cons

Limited information on deployment and serving
Deprecation of pipeline run templates in favor of snapshots

Reviews

No reviews yet. Be the first!

Q&A

Can the MCP server be used for model serving or deployment monitoring?

While the server exposes metadata about deployments, serving snapshots, and service endpoints, detailed deployment and serving management features are limited and not fully covered.

Asked by Youssef El-Sayed · Feb 7, 2026

Is there any limitation regarding pipeline run templates?

Yes, the project notes that pipeline run templates are being deprecated in favor of using snapshots, so future workflows should rely on snapshot‑based triggering.

Asked by Zelda Brandt · Feb 3, 2026

What ZenML entities can I read through the MCP server?

It provides live read access to core ZenML entities including users, stacks, pipeline definitions, execution histories, steps (with code and logs), schedules, artifacts metadata, deployments, serving snapshots, and service endpoints.

Asked by Yaw Owusu · Jan 29, 2026

How does mcp-zenml let me trigger new pipeline runs?

The MCP server can start pipeline executions by sending snapshots or run templates to the ZenML API, allowing you to launch runs directly from MCP clients such as Cursor or Claude Desktop.

Asked by Rania Nasser · Jan 24, 2026