We use AI-assisted tooling on the job and have no problem with you using it here. Being honest about how you used it tells us more than pretending you didn't. There is no penalty for heavy use and no bonus for avoiding it — we're interested in your judgment.
Did you use AI tools on this exercise? (yes / no)
If yes, briefly:
- Which tools? (Copilot, Claude, ChatGPT, Cursor, etc.)
- Where did they help most? (e.g. boilerplate, debugging the flaky test, drafting the CI YAML)
- What did you accept vs. reject or rewrite? (Did the AI suggest anything you decided was wrong or didn't fit? How did you catch it?)
- Anything you verified by hand because you didn't fully trust the output?