MCP_cloudera

MCP_cloudera

2

The Cloudera ML Model Control Protocol (MCP) is a Python-based integration that connects with Cloudera Machine Learning services, offering features such as job management, model deployment, and experiment tracking. It is designed to run as a server, facilitating programmatic interaction through Claude or custom Python applications.

Cloudera ML Model Control Protocol (MCP)

This MCP provides a Python-based integration with Cloudera Machine Learning, enabling interaction with CML services.

Features

  • Upload folders while preserving directory structure
  • Create, list, and delete CML jobs
  • Manage ML models, deployments, and experiments
  • Manage CML applications

Configuration

Set host URL, API key, and optionally the project ID for the MCP.

Running the MCP Server

Run as a server to allow interaction with Cloudera ML through Claude using stdio transport.

Usage

Integrate with Claude Desktop or use directly in Python by importing and initializing the MCP.

Command-line Testing

Test MCP functions like listing jobs or uploading folders via a command-line script.