visual-ralph

Transform approved visual references into pixel-verified UI code iterations.

2|Updated May 11, 2026
One-click install
npx skills add https://github.com/xz1220/oh-my-kimi --skill visual-ralph-xz1220
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: visual-ralph
Source: https://github.com/xz1220/oh-my-kimi/tree/main/skills/visual-ralph
Command: npx skills add https://github.com/xz1220/oh-my-kimi --skill visual-ralph-xz1220

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Visual Ralph turns an intended frontend UI reference into a measurable, fail/pass-driven implementation loop instead of subjective visual tweaking, so teams can reliably reach pixel-level match.

Core Features & Use Cases

  • Visual evidence-driven iteration: Uses Visual Ralph verdict output and pixel diff as measurable guidance for each editing cycle.
  • Reference-first workflow: Builds a visual truth source from an approved mockup, static reference, or a URL-derived baseline (without routing new URL cloning through the deprecated $web-clone).
  • Design-system extraction: Ensures the outcome produces reusable tokens/components (colors, spacing, typography, radii, elevation, variants) rather than one-off screenshots.
  • Guardrailed completion criteria: Enforces screenshot commands, viewport/state recording, verdict thresholds (score >= 90), and repo build/lint/test checks.

Quick Start

Ask the assistant to run the visual-ralph workflow to implement your approved UI reference and iteratively verify it with Visual Ralph verdict and pixel-diff until it matches (score >= 90), then extract reusable design tokens into the repository.

Frequently Asked Questions about visual-ralph

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

FAQPage Schema
How do I verify my frontend UI implementation matches the original design mockup?

To verify frontend UI implementation against a design mockup, use screenshot testing and pixel diff diagnostics to generate measurable pass/fail verdicts. This enforces screenshot commands and viewport recording, iterating frontend build cycles until visual evidence confirms a pixel-level match.

What is pixel diff based UI verification and how does it guide frontend code iteration?

Pixel diff based UI verification compares a rendered frontend build against an approved visual reference, outputting a verdict JSON with pixel differences. This measurable evidence guides iterative code editing loops, replacing subjective visual tweaking with auditable fail/pass thresholds.

How do I extract reusable design tokens after completing a visual UI build?

To extract reusable design tokens after a visual UI build, validate the repository through lint and test checks, then parse the implemented components for colors, spacing, typography, radii, and elevation variants to generate a structured design system output.

Can I use a URL-derived baseline as a visual reference for frontend replication?

Yes, you can use a URL-derived baseline as an approved visual reference for frontend UI replication workflows. The implementation loop uses this baseline to generate pixel-verified code iterations and screenshot-based verdict thresholds without routing through deprecated URL cloning tools.

What score threshold is required to pass a visual verification frontend build cycle?

A score greater than or equal to 90 is required to pass visual verification frontend build cycles. This verdict threshold ensures the UI implementation sufficiently matches the visual reference before allowing extraction of reusable design system tokens and components.