ui-to-code

Converts UI mockups, screenshots, and wireframes into project-conformant code implementations.

Updated Jul 12, 2026
One-click install
npx skills add https://github.com/sumitake/agent-collab --skill ui-to-code-sumitake
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ui-to-code
Source: https://github.com/sumitake/agent-collab/tree/main/plugins/agent-collab/skills/ui-to-code
Command: npx skills add https://github.com/sumitake/agent-collab --skill ui-to-code-sumitake

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Turning a visual design into working code requires careful visual reading, spec extraction, and alignment with an existing codebase's conventions, and mistakes in any step produce implementations that drift from both the design and the project. ## Core Features & Use Cases - Visual Spec Extraction: Reads mockups, screenshots, or wireframes with magnification of detail-dense regions, then extracts layout regions, elements, text, spacing, colors, and interaction hints. - Project-Aware Implementation: Inspects the target project first and reuses its framework, components, design tokens, naming conventions, and accessibility patterns instead of inventing new ones. - Honest Verification Boundaries: States explicitly that visual extraction is primary-only and never claims independent cross-family verification that the runtime does not support. - Use Case: A user pastes a dashboard screenshot and asks to build it; the Skill extracts a working spec, implements it with the project's existing component library and tokens, then compares the rendered result against the source image. ## Quick Start Turn this attached UI screenshot into code using my project's existing components and design tokens.

Frequently Asked Questions about ui-to-code

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I turn a UI screenshot into code?▼

Provide the screenshot and ask to implement it in your project. The Skill reads the image with magnification for dense regions, extracts a layout and styling spec, then implements it using your project's existing framework, components, and design tokens.

How to implement a mockup without breaking project conventions?▼

The Skill inspects the project before writing code and reuses its framework, component library, naming conventions, tests, and accessibility patterns. It prefers existing design tokens over hardcoded values and maps element identifiers to project conventions.

Can a second AI model independently verify the visual extraction?▼

No. The current runtime has no managed cross-family image transport, so visual extraction is performed only by the active primary model. The Skill explicitly reports this limitation and never claims a second independent visual read occurred.

Why does UI extraction fail on dense screenshots?▼

Full-view reads of small text, dense charts, or fine UI detail are unreliable and tend to fail confidently. The Skill mitigates this by magnifying detail-bound regions through native zoom, crop actions, or code-execution crops from the full-resolution source.

Does generated UI code include accessibility features?▼

Yes. The Skill adds semantic HTML, keyboard behavior, focus states, and accessible labels even when the static mockup cannot depict them, following the project's existing accessibility patterns.