Traction scoreGitHub stars can be faked, so popularity alone can be misleading. Traction Score looks for broader signs of recent attention, adoption, and active maintenance.
This collection provides a structured framework for UI/UX designers and developers to audit and polish digital interfaces. It focuses on the technical implementation of design systems, from generating semantic color tokens to ensuring compliance with accessibility standards.
By applying these skills, teams can systematize the review of surface depth, optical alignment, and reading order to create more consistent and inclusive user experiences.
Compatible AgentsThe repository documents support for these agents. The skills may also work with other agents that can load SKILL.md files, but they may need some setup or small changes.
Installs: Complete interfaces plugin sourced from the repository root, covering product-interface skills for typography, colors, layout, accessibility, UI polish, and writing · Claude Code
In a Claude Code agent session, enter these commands in order to add the marketplace and then install its interfaces plugin.
Optional voice, content-style, or localization guide
Use the better-writing skill to review the product copy in [path to one screen or flow]. Compare it with nearby copy and any supplied voice or localization guide. Check terminology, button labels, errors, links, capitalization, settings labels, and empty states. Return the skill's severity-ordered findings table with file-and-line locations, proposed wording, rationale, verification, and its Block or Approve verdict. Do not modify files.
Use the better-layout skill to review [path to one screen or component] at the supported viewport sizes I provide. Check grouping, shared alignment edges, reading order, progressive-disclosure cues, control spacing, clipping, localization growth, and RTL-safe logical properties. Return the skill's severity-ordered findings table, mark checks you could not run as Not verified, and finish with its Block or Approve verdict. Do not modify files.
Use the break skill on the existing component at [component path]. Read its props, slots, states, and rendered data; list the applicable stress scenarios and the axes you excluded. Then create one temporary in-app harness page that imports the unchanged component and renders each scenario in a labeled single column using fixture data only. Load it once with the project's existing browser or preview setup and report only visible breakages; do not publish the page or connect it to live production data.