Thread-safety claims in comments
A teammate ran an AI over a gnarly Java concurrency utility and bulk-added explanatory comments, including ones like '// thread-safe: guarded by the lock above' and '// safe to call without synchronization'. The comments make the code feel approachable. Why should claims like these in AI-generated comments worry you more than ordinary comments, and how do you verify them before they mislead the next maintainer?
find_unguarded_fields(access_lines: list[str]) → list[str][["count|write|stateLock","count|read|stateLock"]]out[][["count|write|stateLock","count|read|none"]]out["count"][["total|write|lockA","total|read|lockB"]]out["total"]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 & agentic: describe the solution in plain language (or narrate it) and the coach grades your approach.