what is really nice about this one is that:

a) don't overestimate LLM b) cache the right answer so you don't call the LLM api everytime c) fast enough

idk if we can say "fast", maybe more of a time saver. the total spent analysing UI and writing tests will be superior of writing "playsmart" tests + runtime.

agreed. one may also say cheaper than having one full time employee writing tests and updating them.

- Yes it is a fork. - Because it extended something well established and did not reinvent the wheel. Stopped the forever freeze, answered the top pains encountered in Requests while keeping top level api as before. Genuinely not just an upgrade, but a real QoL improvement.