I have a question. Where exactly is the line drawn when it comes to cheating in a contest? Can LLMs be used just to debug code or fix syntax errors during a contest, or is the use of LLMs/LRMs completely prohibited?

Obviously, generating logic or full solutions using them is totally wrong — that’s not what I’m asking. My question is only about syntax and debugging.

If even syntax correction is not allowed, then by the same logic, shouldn’t Googling syntax during a contest also be considered cheating?