What problem does it solve? Adopting Specification-Driven Development on a legacy or existing project requires writing PRD, architecture, and UX documents from scratch, which is time-consuming when the knowledge only exists in the code. ## Core Features & Use Cases - Codebase Reconnaissance: Systematically inspects README, build manifests, CI/CD configs, and source structure to derive project type, tech stack, entry points, and UI presence. - Gap-Filling Interview: Asks only the questions that cannot be answered from the code, in a single consolidated round. - Document Generation Pipeline: Dispatches subagents to generate prd.md, architecture.md, and conditionally ux.md plus an HTML prototype, then initializes sdd-tracker.yml. - Use Case: You inherit a brownfield web app with no documentation. Run this skill to reverse-engineer the PRD and architecture docs from the source code and a short interview, so the full SDD workflow can proceed. ## Quick Start Ask the AI to initialize SDD on this existing project by inspecting the codebase and generating the shared SDD documents. ## Quick Start Use sdd-init to bootstrap SDD documentation on my existing codebase.