visual-verdict

Compare generated UI screenshots with reference images to produce a deterministic JSON verdict.

Updated Mar 21, 2026
One-click install
npx skills add https://github.com/gtpgg1013/claude-skills-collection --skill visual-verdict-gtpgg1013
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: visual-verdict
Source: https://github.com/gtpgg1013/claude-skills-collection/tree/main/skills/agents/visual-verdict
Command: npx skills add https://github.com/gtpgg1013/claude-skills-collection --skill visual-verdict-gtpgg1013

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Enable deterministic evaluation of visual UI fidelity by comparing a generated screenshot to reference images and returning a strict JSON verdict that guides the next iteration.

Core Features & Use Cases

  • Deterministic pass/fail assessments for UI screenshots against reference images
  • JSON-driven outputs suitable for automation in design QA and regression testing
  • Use Case: When validating layout, spacing, and typography across UI screens during design reviews and rework cycles.

Quick Start

Compare your latest UI screenshot with one or more reference images to obtain a structured JSON verdict for the next edit.

Frequently Asked Questions about visual-verdict

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

FAQPage Schema
How do I automate visual regression testing for UI screenshots against reference images?

Visual regression testing for UI screenshots is automated by comparing generated screenshots against reference images to produce a deterministic JSON verdict. This structured output includes a score, pass/fail status, category match status, and specific differences to guide the next UI iteration.

What is a deterministic JSON verdict in UI evaluation and how does it work?

A deterministic JSON verdict in UI evaluation is a strict, structured output generated by comparing a current UI screenshot to reference images. It removes subjective human judgment by returning a score, pass/fail assessment, category match status, and actionable suggestions for design rework.

Can I use screenshot comparison for design QA during UI iteration cycles?

Screenshot comparison supports design QA during UI iteration cycles by validating layout, spacing, and typography against reference renders. It outputs a structured JSON verdict containing a score, category match status, and specific differences to directly inform necessary design edits.

What inputs are required to perform a UI evaluation with screenshot comparison?

Performing a UI evaluation requires inputs of one or more reference images and a single generated screenshot. The tool compares these inputs to evaluate visual fidelity and outputs a strict JSON verdict detailing differences, suggestions, and a pass/fail score.

Does automated screenshot comparison work for evaluating layout and typography differences?

Automated screenshot comparison works effectively for evaluating layout and typography differences by comparing current renders to reference images. It identifies specific visual discrepancies and returns them within a deterministic JSON verdict to guide corrections and rework.