Reading & Directing AI-Written Code · Basics
Reading & Directing AI-Written Code — Basics
30 practice questions on Reading & Directing AI-Written Code. Every question is written from a specific moment in a real lecture, and after you answer it links to that exact timestamp so you can check it yourself.
What this pack asks
The free questions in this pack. Answer choices and explanations appear as you play.
A helper comes back from an agent, and a one-call unit test passes: def record(msg, seen=[]): seen.append(msg) return len(seen) In the service it is called once per request. What surfaces first under real traffic?
Answer this one →An agent returned this batch gate, and it passes on a hand-made list of good and bad rows: for r in rows: if not r.valid: log("bad row"); break else: publish(rows) Upstream, rows is whatever the last poll returned. What goes wrong first in production?
Answer this one →A hot counting loop written as while i < n: total += i; i += 1 was rewritten by an agent as for i in range(n): total += i, and it benchmarks noticeably faster. The agent's stated reason: the loop's bounds check and increment now happen inside C-level iterator code instead of being executed by Python. Is that reasoning sound?
Answer this one →
Written from lectures by mCoding. Not affiliated with or endorsed by any university, channel, creator or certification program.