Evaluate AI-generated outputs against a defined rubric for accuracy, completeness, reasoning quality, tone, consistency, and policy compliance. Useful for teams developing reusable prompts and AI-assisted workflows.