What problem does it solve? Solo builders using AI assistants often jump straight into coding without verifying that a slice is ready, that the design matches available data, or that the output actually meets its done criteria. This Skill enforces a disciplined build execution workflow so every slice is anchored, verified, and independently reviewed before it is considered complete. ## Core Features & Use Cases - Status-gated slice selection: Only Ready slices enter build, selected automatically by plan priority and journey order, with dependency blocking enforced. - Four-anchor validation: Every slice requires a design anchor, data anchor, done anchor, and process anchor, plus a design-to-data correlation gate that hard-stops on missing fields. - Independent review handoff: The builder runs an observation-only QA pass, then hands off to a Review Agent that independently verifies design fidelity, data store contents, and done criteria before clearing the slice. - Use Case: A solo founder building a web app with Claude Code invokes build mode; the Skill reads the backlog, picks the next Ready slice, confirms the design file and mock data fields match, builds the feature, runs a preview server QA pass, and hands off to the Review Agent for an independent CLEARED or GAPS verdict. ## Quick Start Ask the AI to start the next build slice from the backlog and it will select the highest-priority Ready slice, verify its anchors, and begin building.
