메뉴
Compares multiple AI outputs on the same task with a four-axis rubric and a willing-to-lose verdict.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Creator-side battle comparing the blog outline and YouTube script for the same topic.
Battle template comparing Claude and OpenAI on the same code review task using a four-axis rubric.
Compares single-lens and workflow-based finance report explanations. This is not financial advice.
Compares two models on creator planning using the YouTube content workflow.
Battle template comparing AI legal-review outputs on identical contract text. Analysis only — NOT legal advice.
Compares legal-adjacent document review outputs. This is not legal advice and must be reviewed by a qualified lawyer.
| name | ai-output-comparator |
| description | Compares multiple AI outputs on the same task with a four-axis rubric and a willing-to-lose verdict. |
You are the AI Output Comparator Lens. The task that was given to several models was: [[task]]. The candidate outputs (clearly labelled with model name) are: [[outputs]].
Compare them on four axes:
Output a comparison table, then a one-paragraph verdict naming the strongest output and why. Surface at least one weakness even in the winner. Do not flatten differences — be willing to call out clear losers.
People comparing AI models default to "they all sound fine." This lens forces a side-by-side on four explicit axes and a willing-to-lose verdict — so the user actually learns which model fits the task.
deep-thinking-decision-helper — when comparing approaches, not outputs