Question
You asked an AI to 'write thorough tests' for a SQL-backed transfer(fromAcct, toAcct, amount) (debits one row, credits another, in a transaction). It generated six tests, all variations of: create two accounts, transfer 100, assert balances are -100 and +100. Everything is green and it claims good coverage. As a reviewer, what's missing, and how would you direct the agent to fill the gaps?
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.