behavioral

Installation
SKILL.md

Behavioral design is the practice of applying what psychology knows about human motivation, decision-making, and habit to create products that genuinely earn engagement. The mechanisms here — social proof, commitment, variable reward, goal gradients, visceral appeal — are the same tools the deceptive-patterns skill flags when weaponized. The difference is intent, transparency, and whether the design serves the user's own goals or overrides them.

Load deceptive-patterns as the ethical guardrail when the work involves conversion optimization, sign-up flows, subscription design, or any engagement mechanic where the line between persuasion and manipulation is a live question.

When this applies

  • Understanding why users don't complete a flow — identifying the motivation, ability, or prompt gap using the Fogg behavior model.
  • Designing for habit formation — applying the Hook model's four phases honestly (trigger connected to a genuine user need, variable reward that delivers real value, investment that makes the product more useful over time).
  • Applying Cialdini's principles (reciprocity, commitment/consistency, social proof, authority, liking, scarcity, unity) to a sign-up flow, onboarding, referral mechanic, or trust-building element.
  • Designing for emotional resonance — using Norman's three levels (visceral: immediate sensory response; behavioral: ease and effectiveness of use; reflective: meaning, identity, and story) to shape how a product feels, not just how it functions.
  • Building choice architecture — reducing decision paralysis, sequencing choices, using anchoring and defaults honestly.
  • Retention and re-engagement — understanding when a product earns return visits vs. manufactures compulsion.

Not the dark version of any of these (use deceptive-patterns), not the sequence through time / IA / funnel (use journey), not the visual mood or aesthetic DNA (use core design).

Rules

Installs
4
GitHub Stars
236
First Seen
Jun 21, 2026
behavioral — ryanthedev/design-for-ai