Last released Oct 8, 2026
Evaluate and compare AI agent setups through experiments, inspections, and rubric scoring.