Learn to build an evaluation framework for your agent in seven short modules: the thinking, the types of test, the framework, then running it live. Five of them you do inside Prefactor. Leave with a quality framework for your own agent and the evals to hand to engineering.
Seven modules on building an evaluation framework: defining good, the golden set, types of test, the layered framework, evaluating live, versioning change, and common mistakes. Five include a hands-on activity.
Say what you want to check and get the evals that check it, from Prefactor, the community and open-source projects, ready to copy.
Say whether the agent is still being built or already live, describe it in two fields, and get five evals from the library.
Every run evaluated, every failure shown by category, and what to change.