Treat Your AI's Output Like User Input

A few months ago, I added a small feature to an internal tool. The idea was simple: paste a customer’s support email, and a language model would extract the order ID and draft a reply. It worked on the first try, which honestly should have been my first warning.