ui-judge

Capture URL screenshots at multiple viewports and score UI quality against a five-criterion rubric.

1|Updated Apr 1, 2026

Search Tags:#playwright #ui-testing #screenshot #visual-quality #ui-evaluation #claude-vision

Ready to run

npx skills add https://github.com/charleshall888/cortex-command --skill ui-judge

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: ui-judge
Source: https://github.com/charleshall888/cortex-command/tree/main/plugins/cortex-ui-extras/skills/ui-judge
Command: npx skills add https://github.com/charleshall888/cortex-command --skill ui-judge

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Manual UI quality assessment and visual regression checks are time-consuming and prone to human error. ui-judge automates this process by capturing screenshots from a live URL, evaluating them against a five-criterion rubric, and producing a structured verdict.

Core Features & Use Cases

Capture screenshots of a URL at multiple viewport sizes and score against a five-criterion visual quality rubric.
Run a two-call LLM process per viewport to generate a JSON verdict and identify affected selectors for targeted fixes.
Persist results to ui-check-results/judge.json for single-viewport runs or per-viewport judge-<viewport>.json files for multi-viewport scenarios.

Quick Start

Trigger the skill in Claude Code with /ui-judge to generate a visual quality verdict for the current UI.

Frequently Asked Questions about ui-judge

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

FAQPage Schema

How do I automate UI visual quality evaluation using screenshots?▼

Automate UI visual evaluation by capturing URL screenshots at multiple viewports and scoring them against a five-criterion rubric. A two-step LLM process generates a structured JSON verdict and identifies impacted selectors for targeted fixes.

What is visual regression checking and how does an automated rubric work?▼

Visual regression checking compares UI screenshots against expected quality standards. This skill uses a five-criterion rubric evaluated through a two-call LLM process per screenshot to produce a structured verdict detailing visual quality and affected selectors.

How do I capture screenshots at different viewport sizes for UI testing?▼

Capture screenshots at different viewport sizes for UI testing by triggering an automated evaluation that runs per viewport. Multi-viewport runs generate separate judge-<viewport>.json files, while single-viewport runs write results to judge.json.

Does UI evaluation with Claude vision identify specific CSS selectors to fix?▼

Yes, UI evaluation with Claude vision identifies specific CSS selectors to fix. The two-step LLM process generates a structured JSON verdict and then isolates impacted selectors for targeted UI corrections.

Can I use Playwright screenshots for LLM-based visual quality scoring?▼

Yes, you can use Playwright screenshots for LLM-based visual quality scoring. The skill captures URL screenshots and applies a two-call LLM process to evaluate them against a five-criterion visual quality rubric, outputting a JSON verdict.

What are the limitations of automated UI evaluation and will it fail my build?▼

Automated UI evaluation is advisory only and will not fail your build, as the workflow always exits with code 0. It provides a structured verdict to guide manual fixes rather than enforcing strict pass/fail gates.