What problem does it solve? AI coding agents often jump straight into implementation without understanding requirements, leading to wasted work on unexamined assumptions. This Skill enforces a structured design dialogue that explores intent, constraints, and approaches before any code is written. ## Core Features & Use Cases - Structured Design Dialogue: Explores project context, asks clarifying questions one at a time, proposes 2-3 approaches with trade-offs, and presents the design in sections for incremental user approval. - Spec Document Generation: Writes the approved design to a dated spec file in docs/superpowers/specs/, runs a self-review for placeholders and contradictions, and gates implementation behind explicit user approval. - Visual Companion: Optionally serves browser-based mockups, wireframes, and side-by-side comparisons via a local WebSocket server, recording user click selections for feedback. - Use Case: Before building a new feature, use this Skill to interview yourself about requirements, compare architectural approaches, and produce a committed design document that feeds directly into implementation planning. ## Quick Start Ask the agent to brainstorm the design for your next feature before writing any code, and answer its clarifying questions one at a time.
