visual-verdict

Compare generated screenshots against reference images and return a JSON verdict with score and differences.

Updated Jul 3, 2025
One-click install
npx skills add https://github.com/andrew-t-james/dotfiles --skill visual-verdict-andrew-t-james
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: visual-verdict
Source: https://github.com/andrew-t-james/dotfiles/tree/main/dot_codex/skills/visual-verdict
Command: npx skills add https://github.com/andrew-t-james/dotfiles --skill visual-verdict-andrew-t-james

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Use this skill to compare a generated UI screenshot against reference images and produce a strict JSON verdict that can drive the next edit iteration.

Core Features & Use Cases

  • Structured, deterministic evaluation across layout, spacing, typography, and component styling.
  • Produces a JSON verdict with a numeric score, a pass/revise/fail label, and concrete differences and actionable suggestions.
  • Use cases include QA cycles for UI dashboards, feeds, and app screens to guide design and implementation.

Quick Start

Run the visual-verdict tool to compare a generated screenshot against reference images and produce 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 compare a generated UI screenshot against a reference image for visual differences?

To compare a generated UI screenshot against reference images, the visual-verdict skill identifies visual differences and returns a strict JSON object containing a numeric score, a deterministic verdict, and actionable suggestions to guide the next edit iteration.

Can I automate UI design review and get a deterministic pass or fail verdict?

Yes, you can automate UI design review by running the visual-verdict tool to evaluate layout, spacing, typography, and component styling, which outputs a deterministic pass, revise, or fail label along with a numeric score in a structured JSON format.

What is the best way to detect pixel diffs between app screens during QA cycles?

The best way to detect pixel diffs during QA cycles is using a deterministic screenshot comparison process that evaluates dashboards and feeds, returning a JSON verdict with category matches, concrete differences, and actionable suggestions for implementation.

Does visual regression testing for UI layouts require any external dependencies?

No, visual regression testing for UI layouts using this skill requires no external dependencies, allowing you to directly compare screenshots against reference images and produce a structured JSON verdict without additional environment setup.

What specific UI elements are evaluated when comparing screenshots for visual differences?

When comparing screenshots for visual differences, the evaluation specifically targets UI layout, spacing, typography, and component styling across app screens, feeds, and dashboards to produce a comprehensive numeric score and deterministic verdict.