Judge test of transferFunds adequacy
You asked an AI agent to write a test for `transferFunds(from, to, amount)` in your payments service. It produced a test that mocks the database, calls `transferFunds`, and asserts the function returned `{ success: true }`. The test passes. Judge whether this is a meaningful test of a money-moving operation, and what you'd require instead.
Implement
transfer_violations(from_before_cents: int, to_before_cents: int, from_after_cents: int, to_after_cents: int, amount_cents: int) → list[str]Examples
in
[10000,5000,7500,7500,2500]out[]in
[10000,5000,7500,5000,2500]out["destination_credit_wrong","total_not_conserved"]in
[10000,5000,10000,5000,2500]out["source_debit_wrong","destination_credit_wrong"]What a strong answer looks like
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.
0:00 of about 16 min
Vibe & agentic: describe the solution in plain language (or narrate it) and the coach grades your approach.
Which questions mattered is sealed until you submit. Telling you now would just be handing over the edge cases.
Run or narrate your approach, then ask the coach.