visual-verdict

Analyze UI screenshots against reference images and output structured JSON verdicts.

Updated Apr 20, 2026
One-click install
npx skills add https://github.com/jimmi2051/oh-my-copilot --skill visual-verdict-jimmi2051
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: visual-verdict
Source: https://github.com/jimmi2051/oh-my-copilot/tree/main/plugins/omc-copilot/skills/visual-verdict
Command: npx skills add https://github.com/jimmi2051/oh-my-copilot --skill visual-verdict-jimmi2051

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Structured visual QA verdicts for screenshot-to-reference comparisons.

Core Features & Use Cases

  • Deterministic JSON verdicts that quantify visual differences and provide a clear pass/revise/fail signal.
  • Supports multiple reference images to handle layout variations and real-world UI states.
  • Outputs a strict JSON payload including score, verdict, category_match, differences, suggestions, and reasoning for automated gating.
  • Use Case: Validate a new UI against design references to drive targeted edits and ensure consistency.

Quick Start

Run this skill on a generated screenshot and the reference images to obtain a 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 automate visual regression testing for UI screenshots against design references?

Automate visual regression testing by analyzing a generated UI screenshot against reference images to produce a structured JSON verdict. This output describes visual differences and required edits, enabling automated gating in design QA workflows.

What is a JSON visual verdict for screenshot comparison?

A JSON visual verdict is a structured payload that quantifies visual differences between a UI screenshot and a reference image. It enforces a fixed schema including score, verdict, category_match, differences, suggestions, and reasoning to provide a clear pass, revise, or fail signal.

How do I validate UI layout, typography, and colors across multiple screen captures?

Validate UI layout, typography, and colors by running a screenshot comparison skill that supports multiple reference images. It analyzes these specific styling properties across screen captures and outputs a deterministic JSON verdict detailing any visual differences and required edits.

Can I use screenshot comparison for automated gating in design QA workflows?

Yes, you can use screenshot comparison for automated gating by leveraging the strict JSON payload output. The deterministic score and verdict fields provide a clear pass, revise, or fail signal suitable for integrating automated visual checks into continuous integration pipelines.

Does UI testing with image diff work with multiple reference images for layout variations?

Yes, UI testing with image diff supports multiple reference images to handle layout variations and real-world UI states. This allows the analysis to account for different valid configurations when validating a generated UI against design references.

What's the best way to get a deterministic pass or fail signal for UI visual differences?

The best way to get a deterministic pass or fail signal is to use a structured visual QA tool that enforces a fixed JSON schema. It quantifies visual differences and outputs a clear verdict, ensuring consistent automated validation against design references.