Summary: An intentionally minimal AI evals runner that illustrates how to test LLMs.
Publication date: 2026-07-12