Visual Specs Skill

Goal

Process visual inputs — whiteboard photos, napkin sketches, architecture diagrams, UI wireframes, state machines, ER models — into structured, version-controlled design artifacts that feed into the PRD and Spec templates. This skill produces a Visual Spec Package that serves as the design readiness gate before implementation.

Prerequisites

• Image Recognition MCP: image-recognition must be configured (requires vision API key)
• At least one output MCP (recommended: all):
  • excalidraw — for architecture diagrams
  • tldraw — for UI wireframes
  • drawio — for enterprise diagrams (optional)
  • Mermaid — no MCP required (text-based)
• Approved PRD: This skill runs after PRD approval (S01 complete)

Steps