Intermediate 20-25 minutes
Encode Your Values as Observable Assistant Behaviors
Translate abstract values into concrete response choices, trade-off rules, and examples your AI twin can consistently apply.
Recommended tools: Claude, ChatGPT
Live Preview
ROLE
Act as a values-to-behavior design facilitator.
CONTEXT
My stated values are [VALUES LIST]. Examples of decisions that reflect them are [DECISION EXAMPLES]. The twin will operate in [OPERATING CONTEXT].
OBJECTIVE
Create behavior rules that express my values without reducing them to slogans or claiming certainty about my beliefs.
INSTRUCTIONS
For each value, identify observable actions, communication habits, prohibited shortcuts, and evidence from the examples. Distinguish explicit commitments from your interpretations. Explore situations where two values could conflict, such as speed versus care or candor versus kindness. Propose a transparent tie-breaking process that asks for human guidance when necessary.
Write six short scenarios showing how the twin should respond under pressure. Include at least one case where the correct behavior is to acknowledge uncertainty rather than produce a confident answer. If Christian commitments are supplied, represent them respectfully without assuming denominational positions that were not stated.
CONSTRAINTS
Avoid fabricated statistics and flag assumptions. Do not infer religious, political, or other sensitive beliefs beyond what I explicitly provide.
OUTPUT FORMAT
Return a values matrix, conflict-resolution rules, scenario responses, and clarification questions.
Open in:We copy the prompt first, then open the tool in a new tab. Results vary by tool.