What problem does it solve? Designing AI-powered interfaces raises questions that general UI guidelines do not answer: how users write prompts, how AI output earns trust, how memory and consent are handled, and how humans stay in control of agentic workflows. This Skill supplies a structured catalog of AI-specific interaction patterns so designers and developers can make consistent, evidence-based decisions. ## Core Features & Use Cases - 60+ patterns in six categories: Wayfinders, Prompt Actions, Tuners, Governors, Trust Builders, and Identifiers, each with design guidance and usage criteria. - Selection frameworks: Decision tables mapping user problems and product types (chat assistant, code copilot, AI agent) to the right patterns, plus a trust-level decision tree. - Anti-pattern guardrails: Critical gotchas such as avoiding photorealistic avatars, never hiding memory contents, and always providing stop controls during generation. - Use Case: When building an enterprise document AI, consult the catalog to combine Attachments, Citations, Caveat, and Disclosure patterns so users can verify AI-generated summaries against sources. ## Quick Start Ask the AI to recommend interaction patterns for a chat assistant that needs onboarding suggestions, follow-up prompts, and transparent memory controls.