A test either passes or it does not. An agent gives you a different answer every time, so “it works” stops being a thing you can assert and becomes a thing you have to measure. This lab is about that measurement, and about the disciplined loop that turns it into improvement: diagnose, change one thing, measure, keep it or throw it away. Eleven sections, plain English first, and simulations for the parts that only make sense once you have watched a number move the wrong way.