About
Wraps Snowpark ML so Cortex Code can author feature pipelines, training jobs, and model registry entries from the terminal. Strong defaults around reproducibility — every artifact gets a version, every training run gets a manifest, every deployment is rollback-safe.
How to use
Describe what you want and the skill produces the pipeline:
cortex-code 'train a churn model on customer_events with a 90-day window'The output is a runnable Python file (or notebook), a kickoff command, and a registry entry pre-filled with the metrics from the training run. Promotion to production is a separate, explicit step.
Tags
snowparkpipelinesregistry