Simulate an interview panel as agents that design questions, score independently on anchored rubrics, and reconcile in a debrief. Use when you want to pressure-test a hiring loop, calibrate rubrics, or dry-run questions, not to score a real candidate.
