Question
You ask an AI to write a SQL migration that adds a `NOT NULL` column `status` defaulting to `'active'` to a 200M-row Postgres `orders` table. It writes `ALTER TABLE orders ADD COLUMN status text NOT NULL DEFAULT 'active';`. In staging it's instant. You ship to prod and it takes a full table lock for 9 minutes, taking down checkout. You re-prompt 'make it not lock'; it wraps the same statement in a transaction, which changes nothing. What did the model not know, and how do you re-steer?
Treat the AI’s output as a draft to verify, not an answer to trust. Name the specific flaw and the input that triggers it, say how you’d catch it — tests, edge cases, reading critically — and how you’d re-prompt or decompose to get it right.
Vibe coding: describe the solution in plain language (or narrate it) and the coach grades your approach. Generating runnable code from your description is coming next.