Automated evaluation platform for AI agents that runs tests inside high-fidelity simulation environments. Provides scalable, repeatable testing and benchmarking for AI developers and researchers.