visual-verdict

Compare UI screenshots against reference images and output a deterministic JSON verdict.

Updated Mar 10, 2026
One-click install
npx skills add https://github.com/eszxcvfd/ANIZAKI --skill visual-verdict-eszxcvfd
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: visual-verdict
Source: https://github.com/eszxcvfd/ANIZAKI/tree/main/.agents/skills/visual-verdict
Command: npx skills add https://github.com/eszxcvfd/ANIZAKI --skill visual-verdict-eszxcvfd

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This tool provides automated, deterministic judgments on UI visuals by comparing a generated screenshot to reference images, reducing manual review cycles.

Core Features & Use Cases

  • Pixel-perfect comparison for layout, typography, spacing, and component styling.
  • Supports multiple reference images and optional category hints to guide evaluation.
  • Returns a strict JSON verdict with score, verdict, category_match, differences, suggestions, and reasoning to drive next edits.

Quick Start

Run visual-verdict with your latest UI screenshot and one or more reference images to obtain a strict JSON verdict.

Frequently Asked Questions about visual-verdict

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

FAQPage Schema
How do I perform a visual diff between a UI screenshot and reference images?

To perform a visual diff, compare your generated UI screenshot against single or multiple reference images. This process evaluates layouts, typography, spacing, and component styling to produce a deterministic JSON verdict.

What is a deterministic visual QA verdict for UI components?

A deterministic visual QA verdict is a strict JSON output containing score, verdict, category_match, differences, suggestions, and reasoning fields, generated by comparing a screenshot to reference images.

Can I use a category hint to guide UI testing and pixel diff evaluation?

Yes, you can provide an optional category hint alongside your UI screenshot and reference images to guide the pixel diff evaluation, ensuring the assessment focuses on relevant component styling and layout criteria.

What is the best way to automate UI review for layout and typography spacing?

Automate UI review by running a visual diff comparison between your latest screenshot and reference images. This generates a strict JSON verdict with specific suggestions and reasoning to drive next edits.

Does visual QA support multiple reference images for a single UI screenshot?

Yes, visual QA supports comparing a single generated UI screenshot against multiple reference images, producing a deterministic JSON verdict that details layout, typography, and spacing differences.