Multi-Perspective Behavior Judge

Problem

Behavior reviews often skew when driven by a single value system: rules, empathy, utility, and intent can produce different verdicts. skill-multi-judge makes that disagreement explicit by scoring the same everyday event from five fixed personas, helping engineers and users see where the disagreement comes from instead of receiving a one-word judgment.

How It Works

• Trigger: Provide judge + a specific personal behavior event, for example judge cutting in line to buy goods because of time pressure.
• Persona dimensions: Rational rules, empathetic feeling, traditional morality, realistic utility, and idealist tolerance each produce a 0–10 score and a one-sentence rationale.
• Weighted aggregation: Scores are combined using 25%, 25%, 20%, 15%, and 15%, then mapped to bands such as excellent, basically correct, neutral, clearly wrong, or seriously inappropriate.
• Output structure: The response consistently includes individual persona scores, weighted total score, final evaluation, and two actionable improvement suggestions.

Scope and Limits

It is intended for reflecting on everyday personal behavior, motives, consequences, and wording, not for illegal activity, violence, personal attacks, sensitive political issues, or extreme ethical cases; such inputs are rejected. If the event lacks detail, the rationale may remain generic, so adding context, affected parties, frequency, and outcomes improves the result.