Claude, or any LLM, can be used every day for product design work. Not only as a chatbot you ask questions to, but as something closer to a copilot who shares the same references you do. Those references slowly turned into skills: one on the Laws of UX, one on debunked UX myths.
A skill is a folder Claude reads before it answers. Not a prompt retyped each time, a standing reference it consults on its own when the conversation matches. The ux-laws skill grounds a critique in Fitts, Hick, Jakob, Gestalt, the usual names, instead of "this feels off." The uxmyths skill catches the reverse: a design decision resting on something that sounds true and isn't. "More choices make users happier." It doesn't.
Writing them down forced a distinction I hadn't made explicit before: knowing a framework and having it ready at the right moment are different skills. I know Hick's Law but I don't always reach for it during design review, or when a dropdown has grown to forty items. Having the skill ready makes it easier to stay pragmatic, and to argue from references instead of feel.
Using them daily, and having colleagues push back on the arguments they produced, settled it: everyone designing deserves to work from the same references. More skills are coming, each one a shortcut back to the research instead of a rule pulled from memory.
I’m Maxime Boulogne, I share these notes and case studies to prove UX only serves the light side of the Force.