Visual QA utilizes vision models to review screenshots against design intent, identifying issues such as spacing, alignment, and accessibility. This skill is aimed at designers and developers who want to ensure their work aligns with established design standards before deployment.
$ npx skills add https://github.com/dylanfeltus/skills --skill visual-qaVisual QA leverages vision-capable AI models to automatically review screenshots against design standards. The skill evaluates seven key categories—layout, typography, color, hierarchy, components, polish, and responsive behavior—providing structured feedback on spacing, alignment, color consistency, and accessibility gaps. It supports full reviews, quick reviews, and mockup comparisons, helping designers and developers catch issues before shipping. Integration with build workflows enables QA checks during development. The skill requires a vision-capable model like GPT-4o or Claude and the ability to capture screenshots via browser tools or user-provided images.
Install the skill using the provided npm command.
User requests to review a design or screenshot.
After building a page, verify design accuracy.
Compare implementation against mockups.
User asks if the design meets expectations.
$ npx skills add https://github.com/dylanfeltus/skills --skill visual-qagit clone https://github.com/dylanfeltus/skillsCopy the install command above and run it in your terminal.
Launch Claude Code, Cursor, or your preferred AI coding agent.
Use the prompt template or examples below to test the skill.
Adapt the skill to your specific use case and workflow.
Check the GitHub repository or documentation for usage examples.
Take a free 3-minute scan and get personalized AI skill recommendations.
Take free scan