I turned Andrej Karpathy’s tips on understanding LLM output into an Open Source Agent Skill : [make-it-click](https://github.com/ajithraghavan/make-it-click)Inspired by his X post : as Models do more of the work, we spend more time understanding it. So the fo…