zenml-io/mcp-zenml

Overview

An MCP server that acts as a bridge between MCP-compatible clients (such as Cursor and Claude Desktop) and ZenML MLOps and LLMOps pipelines. It enables these clients to interact with pipeline runs, artifacts, and deployments through the Model Context Protocol.

Best for

Best for
ZenML users who want to interact with their MLOps and LLMOps pipelines through MCP-enabled tools and AI assistants

Use cases

• Query pipeline status and logs from an AI assistant inside a code editor
• Trigger ZenML pipeline runs or deployments via natural language commands
• Retrieve artifact metadata and model versions through a chat interface

How to use

Install

npx cloudflared tunnel --url http://localhost:8001

Tools exposed

• get_snapshot
• list_snapshots
• get_deployment
• list_deployments
• get_deployment_logs
• trigger_pipeline
• get_active_project
• get_project
• list_projects
• get_tag
• list_tags
• get_build
• list_builds
• list_artifacts
• list_secrets
• open_pipeline_run_dashboard
• open_run_activity_chart
• stack_components_analysis
• recent_runs_analysis
• most_recent_runs

Tested with

Claude Desktop, Claude Code, Cursor, VS Code, ChatGPT

Notes

An MCP server that acts as a bridge between MCP-compatible clients (such as Cursor and Claude Desktop) and ZenML MLOps and LLMOps pipelines. It enables these clients to interact with pipeline runs, artifacts, and deployments through the Model Context Protocol.

45 stars on GitHub. Last updated 2026-06-01. Licensed MIT.

Use cases

• Query pipeline status and logs from an AI assistant inside a code editor
• Trigger ZenML pipeline runs or deployments via natural language commands
• Retrieve artifact metadata and model versions through a chat interface

Pros

• Native integration with ZenML’s MLOps and LLMOps workflows
• Works with popular MCP clients like Cursor and Claude Desktop
• Open source and written in Python, easy to extend or customize

Cons

• Low GitHub star count (45) suggests limited adoption or early development stage
• Requires a running ZenML server and configured pipelines to be useful
• Documentation and community support may be sparse compared to larger projects

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.