BOOTOSHI (@KingBootoshi)
AI(예: Codex, Claude)에 피드백을 줄 때 단순히 오류만 지적하지 말고 '무엇을 잘했는지'와 '무엇을 잘못했는지'를 모두 명확히 알려주라는 프롬프트/피드백 요령을 제시. 이 간단한 방법이 출력 품질을 보장하도록 도와준다는 실무 팁이다.

BOOTOSHI 👑 (@KingBootoshi) on X
man this is such a simple ai/codex/claude code hack that everyone seems to forget for a GUARANTEED better output when you're giving feedback you need to tell the ai what you DID LIKE and what you DID NOT LIKE tell it what it did RIGHT and what it did WRONG if you only say

