Question
An AI agent generated setup docs for a TypeScript monorepo. You followed them on a clean machine and step 4 fails — `pnpm build` errors because the docs never mention you must copy `.env.example` to `.env` and set two required keys first. The AI couldn't have known this from the code alone. How do you produce trustworthy setup docs with an AI given that some required steps aren't visible in the source?
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.