Verified:GitHub

visual-verdict

Evaluate UI screenshots against reference images and produce a structured JSON verdict with score, pass/fail status, category match, differences, suggestions, and reasoning.

Updated Apr 10, 2026

Search Tags:#ui #qa #json #comparison #screenshot #visual

Ready to run

npx skills add https://github.com/lpjhelder/kronon-forge --skill visual-verdict-lpjhelder

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: visual-verdict
Source: https://github.com/lpjhelder/kronon-forge/tree/main/skills/visual-verdict
Command: npx skills add https://github.com/lpjhelder/kronon-forge --skill visual-verdict-lpjhelder

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Visual fidelity checks for UI designs often require manual, time‑consuming comparisons between generated screenshots and reference images, slowing down iteration cycles.

Core Features & Use Cases

Structured JSON verdict: Returns a score, pass/fail status, category match, differences, suggestions, and reasoning.
Threshold‑based looping: Stops editing only when the score meets the 90+ pass threshold.
Use case: Ideal for developers and designers who generate UI screenshots and need deterministic feedback before proceeding to code changes.

Quick Start

Use the visual‑verdict skill to compare your UI screenshot with reference images and receive a JSON verdict indicating pass or fail and suggested edits.

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 QA for UI screenshots against reference designs?▼

Automate visual QA by comparing UI screenshots against reference images to produce a structured JSON verdict. This verdict includes a score, pass/fail status, category match, differences, and suggestions to guide automated editing loops.

What is a structured JSON verdict for visual fidelity checks?▼

A structured JSON verdict for visual fidelity checks is a data response containing a score, pass/fail status, category match, differences, suggestions, and reasoning. It provides deterministic feedback for UI design reviews.

How do I use a score threshold to stop automated UI editing loops?▼

Use a score threshold to stop automated UI editing loops by evaluating the JSON verdict from a visual fidelity check. The loop continues editing until the screenshot meets a 90+ pass threshold.

Can I compare one generated screenshot against multiple reference images?▼

Yes, you can compare one generated screenshot against one or more reference images. The visual QA process evaluates the screenshot and returns a JSON verdict detailing differences and category matches.

Do I need any external dependencies to run visual fidelity checks on UI designs?▼

No external dependencies are required to run visual fidelity checks on UI designs. The skill operates independently to evaluate screenshots and generate a structured JSON verdict.