You don’t need to over-engineer for LLMs. If your API or CLI works for humans, it works for AI. TurboPuffer CEO @sirupsen explains: “I think it’s a little overblown how much you have to design for this. It’s the same with humans - we fat-finger stuff.” “The LLMs are a lot more resilient now than they were even just six months ago.” “To me, it’s always about good API design. If you’re designing a good API, a good CLI, or a good interface, you should be able to predict - if I do this, it’s probably going to work this way.”