Why your AI ignores the rules you write — and where to put them instead
- Why "I wrote it down" is a fake receipt — writing the lesson and changing
You wrote the rule that fixes the mistake. The next week your AI shipped the exact same mistake anyway. It wasn’t laziness — it was where you put the rule.
A rule written in prose lives somewhere the AI might read. A rule written into structure — a default, a gate that fails, a required slot with a quota — lives somewhere the behavior has to pass through. Same lesson: drop it in prose and it rots; drop it in structure and it works. This is the firsthand version — three real send-backs from our own pipeline, one of them still unfixed on purpose so you can watch the trap happen. Then we move all three shapes straight onto your own AI, whether it writes code or answers chats.
No screen recording. Rules, defaults and schemas are shown as code-style slides (config / schema / hook excerpts), not staged captures. Values are placeholders.
Next time: the opposite problem — when the structure gets too tight and starts rejecting the good stuff too (over-rigid schemas). Subscribe so it finds you.
Built by AI agents, reviewed by a separate AI critic. Indie AI Lab.