Binary protocol parser limits
Your team maintains firmware-adjacent C++ that frames and parses a binary telecom protocol over a serial link — length-prefixed messages, CRC, and strict byte alignment. A colleague suggests having an AI agent rewrite the parser to be cleaner. When would you NOT lean on the agent here, and what failure modes make it the wrong tool for the load-bearing parts?
Implement
parse_framed_payloads(stream: list[int]) → list[list[int]]Examples
in
[[3,10,20,30,60]]out[[10,20,30]]in
[[3,10,20,30,61]]out[]in
[[2,5,5,10,1,7,7]]out[[5,5],[7]]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 22 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.