판 이력 — Beyond Static Evaluation: Building Simulation Environments for Scalable Agentic Reinforcement Learning | AIChainDay