{"type":"mcp_server","name":"mcp-zenml","description":"This repository implements an MCP server for ZenML, enabling access to live information about ZenML resources like pipelines, artifacts, and users, and allowing triggering new pipeline runs. This repository implements","category":"AI","language":"Python","stars":47,"forks":3,"owner":"zenml-io","github_url":"https://github.com/zenml-io/mcp-zenml","homepage":"https://www.zenml.io","setup":"## Setup\n\n1.  Have access to a ZenML Cloud server.\n2.  Install `uv` locally, following the [`uv` documentation](https://docs.astral.sh/uv/getting-started/installation/).\n3.  Clone the repository locally: `git clone https://github.com/zenml-io/mcp-zenml.git`.\n4.  Create an MCP config file (JSON) with the following structure:\n\n   ```json\n   {\n       \"mcpServers\": {\n           \"zenml\": {\n               \"command\": \"/usr/local/bin/uv\",\n               \"args\": [\"run\", \"path/to/zenml_server.py\"],\n               \"env\": {\n                   \"LOGLEVEL\": \"INFO\",\n                   \"NO_COLOR\": \"1\",\n                   \"PYTHONUNBUFFERED\": \"1\",\n                   \"PYTHONIOENCODING\": \"UTF-8\",\n                   \"ZENML_STORE_URL\": \"https://your-zenml-server-goes-here.com\",\n                   \"ZENML_STORE_API_KEY\": \"your-api-key-here\"\n               }\n           }\n       }\n   }\n   ```\n\n5.  Replace the dummy values in the config file with the correct paths and credentials.\n6.  For Claude Desktop:\n    - Install Claude Desktop.\n    - Open the 'Settings' menu, click on the 'Developer' tab, and click 'Edit Config'.\n    - Paste the config file content into the opened JSON file.\n    - Restart Claude Desktop.\n7.  Optionally, improve ZenML tool output display in Claude by adding a preference in Settings -> Profile.\n8.  For Cursor:\n    - Install Cursor.\n    - Create a `.cursor` folder in the root of your repository.\n    - Inside it, create a `mcp.json` file with the config file content.\n    - Enable the ZenML server in Cursor settings.","tools":"## Available Tools\n\n1.  Users (access live information about users).\n2.  Stacks (access live information about stacks).\n3.  Pipelines (access live information about pipelines).\n4.  Pipeline runs (access live information about pipeline runs).\n5.  Pipeline steps (access live information about pipeline steps).\n6.  Services (access live information about services).\n7.  Stack components (access live information about stack components).\n8.  Flavors (access live information about flavors).\n9.  Pipeline run templates (access live information about pipeline run templates).\n10. Schedules (access live information about schedules).\n11. Artifacts (access metadata about data artifacts).\n12. Service Connectors (access live information about service connectors).\n13. Step code (access live information about step code).\n14. Step logs (access step logs if the step was run on a cloud-based stack).\n15. Trigger new pipeline runs (if a run template is present).","faq":null,"created_at":"2025-02-22T15:49:15+00:00","updated_at":"2025-03-26T06:14:16+00:00","source_url":"https://model-context-protocol.com/servers/zenml-mcp-server-mlops-llmops-pipelines","related_articles":[]}