Batch docstrings vs interactive explain
You need to generate documentation strings for ~8,000 internal functions overnight — no human is waiting, results land in a PR people review in the morning. Separately, you maintain an interactive 'explain this function' button in the IDE. A teammate uses identical model settings (same tier, same low max-tokens, streaming on) for both. What would you tune differently between these two jobs, and why?
Implement
plan_inference_config(human_waiting: bool, call_count: int, max_output_tokens: int) → list[str]Examples
in
[false,8000,400]out["mode=batch","streaming=off","max_output_tokens=1024"]in
[true,1,400]out["mode=sync","streaming=on","max_output_tokens=400"]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 15 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.