Eval
Also called Evaluation
A test for an AI system: give it a task, grade the result, and do that many times, because it may answer differently each time. The result is a rate, not a pass or a fail.
It is how you test the agent's setup and not its code: did that AGENTS.md change help, or did it only feel like it? Without evals you are judging your harness from a few memorable runs.
Say it to your agent
“Before you change the system prompt, build an eval: these 20 tasks, 5 runs each, graded by the hidden scenarios in acceptance/. Run it on the old prompt and the new one, and show me both pass rates with their ranges.”