scarf

Runs David Rock's SCARF model to read a situation for social threat and reward across five domains: Status, Certainty, Autonomy, Relatedness, and Fairness. The brain treats social experience much like physical survival: threat in any domain triggers an avoid response (defensiveness, disengagement), reward triggers an approach response (engagement), and a single high threat outweighs rewards spread across the rest. The output is specific moves that lower threat and raise reward, not labels for feelings.

The five domains, their threat and reward triggers, the practical levers, and the pitfalls live in references/scarf.md. Load that file before analyzing. SCARF is a popularization of neuroscience, a useful heuristic, not a measurement instrument; keep that in view throughout.

When to use

The user is about to do something to or with people who may feel threatened: announce a change, give feedback, send a difficult message, reorganize a team, or make a decision others will react to. The unit of analysis is one situation and the people it affects, read for social threat. Use the change cluster (kotter-change, adkar, force-field-analysis) to plan the change itself and psychological-safety for the standing climate of a team; reach for SCARF when the question is specifically where will this land as a threat, and how do I defuse it.

Language

Conduct the session and write the report in the language the user is writing in. The reference file is English; keep the framework keyed to it regardless of the output language.

Inputs

• SITUATION (required): the change, message, decision, or interaction to analyze. Push for one concrete situation. If it is vague, narrow it with a question or two first.
• PEOPLE (optional but valuable): who is affected and from whose point of view to read the threat (a team, a peer, a report, a whole org). Different people feel threat in different domains; name them where you can.
• CONTEXT (optional): history, prior reactions, constraints, and what the user wants the interaction to achieve. This shapes which domains matter most and how the moves are framed.