Last released Apr 13, 2026
Agent-as-judge grading framework for evaluating AI agent outputs against rubric criteria
Supported by