Open-source LLM testing and evaluation framework. Test prompts, compare models, catch regressions, and red-team your AI applications with automated evaluation pipelines for production-grade reliability.