What problem does it solve? Turning a UI screenshot or product idea into working frontend code normally requires manual decomposition, asset creation, and repeated visual comparison. This Skill automates that pipeline by routing every request into one of three workflows (recreate, redesign, create), generating bitmap assets through configured image channels, and verifying rendered output against the source of truth. ## Core Features & Use Cases - Three workflow modes: Recreate faithfully rebuilds a screenshot against the original reference; Redesign and Create generate an Effect Image first, then implement and verify against it. - Durable runtime with resume: The image2-ui CLI runs, inspects, and resumes executions with persisted state, bounded verify/fix loops, and optional multi-agent DAG scheduling for review, accessibility, and QA. - Asset and icon discipline: Code owns text, controls, and icons from approved libraries, while the image2 wrapper generates photos, cutouts, and illustrations with provenance records and UI-glyph exclusion rules. - Use Case: Provide a mobile app screenshot and ask for a clickable recreation; the Skill decomposes the layout, implements real code-rendered controls, generates needed imagery, and runs image2-ui validate against the reference until findings are resolved. ## Quick Start Ask the assistant to recreate the attached UI screenshot as a clickable, editable frontend and verify it against the original reference.
