Devotd Github
NewCoding agents

ngocanhnckh/renderlint — Visual Defect Checks on Browser Screens

ngocanhnckh/renderlint is a JavaScript tool that inspects web screens in a real browser to find overflow, low contrast, element overlap, and responsive errors.

Repositoryngocanhnckh/renderlint ↗

ngocanhnckh/renderlint is a JavaScript tool that inspects web screens in a real browser to find overflow, low contrast, element overlap, and responsive errors.

GitHub description: Lint the rendered UI: catches overflowing text, low contrast, overlaps, broken responsive layouts, emoji icons and more in any web framework. Built for AI coding agents.

stars
★ 105
Forks
23
Language
JavaScript
License
MIT
Read original source ↗View on GitHub ↗

ngocanhnckh/renderlint is an inspection tool that launches a web app in a real browser and looks for visual defects on the screen. It checks for text overflowing outside cards, low text contrast, elements overlapping one another, and layouts that break at certain screen widths. The repository description says it can be used with any web framework and is designed to inspect screens produced by coding agents. No new version or release information is given. The repository language is JavaScript, and it is published under the MIT license.

The key point is that it checks the screen the browser actually rendered rather than only reading the source code. The documentation says it inspects multiple screen sizes and dark mode, and reports the location of each issue, a screenshot of the screen, and a suggested direction for the fix. It also explains how to attach a logged-in state to the inspection. Methods for integrating with browser automation and Storybook inspections are included in the documentation, so a flow that inspects UI components or the screens shown after login can be set up. This is the scope of the features described in the documentation, and no guarantee is given that it behaves the same way in every app.

The tool's use is limited to finding defects that are visible on the rendered screen. The provided information contains no measurements of detection accuracy or processing time. It can therefore be used to identify what to fix based on the screenshots and issue locations, but there is not enough basis to treat it as an overall evaluation of screen quality or as a replacement for other verification.