AI Prompts

Prove a dbt incremental model handles updates, late data, deletes, and full refreshes, with fixtures that would catch silent drift.

Build a layered test suite that catches transformation, contract, and replay defects before production.

Define practical data-quality checks and thresholds for a critical table, including fail vs warn severity and owner next steps.

Audit a pull request for production risk, missing tests, observability gaps, and a rollback plan instead of style-only review comments.

Design a test strategy for a risky change with the cases that would have caught the last incident, not a generic unit-test shopping list.