Cortex provides a step-by-step playbook for building a production-ready ML pipeline. It helps the agent detect the project's ML stack, define success metrics, create a simple baseline, implement data validation and feature pipelines, train and evaluate models, and deploy a serving endpoint with monitoring. The guidance emphasizes reproducibility, experiment tracking, and avoiding premature complexity.
Use Cortex when a user asks the agent to build or prototype a machine learning model, set up a training/evaluation loop, or create a prediction-serving pipeline. Ideal for classification, regression, and structured-data problems where a clear metric and baseline exist.
Likely compatible with general-purpose coding assistants and agents that can run and edit code (Copilot/Codex, Claude Code, Cursor). The skill assumes filesystem access and ability to run shell commands for file inspection and script execution.
This skill has not been reviewed by our automated audit pipeline yet.
