Entendi has shut down.

Entendi was about keeping "vibe-coding" honest with what AI built for you. If a model writes your code, you should still understand it well enough to defend it. So at each turn we probed that understanding and flagged the knowledge gaps.

We think the problem is bigger: the more capable these models get, the easier it becomes to let them do your thinking. Some of us are going to focus on AI safety, others on how to actively keep people sharp and research that.

Thanks to everyone who tried it, gave feedback, and followed along!