What problem does it solve? Turning a screenshot, wireframe, or HTML mockup into an implementable React component architecture is error-prone: teams mislabel containers, duplicate components across routes, and lose track of which dialogs and forms to build. This Skill converts a visual artifact into a rigorous ## UI Composition spec section that downstream build skills can execute directly. ## Core Features & Use Cases - Citizen Classification: Maps every visible boundary to exactly one of six roles (Route Shell, Section, Component, Leaf, Dialog, Form) using an explicit taxonomy and visual cue catalog. - Structured Output: Produces seven fixed sub-sections — URL contract, ASCII layout map, component tree, per-component anatomy with ASCII mockups, data cards, reuse decisions, and an ordered hand-off table. - Reuse Governance: Searches primitive, shared, and route-local tiers to decide reuse vs. promote-to-shared vs. create-new, citing existing file paths. - Use Case: Given a screenshot of a patients list page, produce the full composition plan — StatsSection with StatCard leaves, PatientListSection owning the list query, and a hand-off table ordering /route, /component, and /form invocations. ## Quick Start Analyze this screen mockup and generate the UI Composition section for the spec, classifying each region and listing the components to build.