jakubkrehel/skills

A collection of agent skills that help you build great interfaces.

What it solves

Building high-quality user interfaces is complex, requiring constant attention to accessibility, typography, layout, and visual polish. This project provides a collection of specialized agent skills that automate these design engineering tasks, helping developers identify regressions and build more consistent, professional interfaces.

How it works

The project functions as a set of modular skills that can be integrated into development environments like Claude Code. These skills act as specialized assistants that can review code, generate design tokens, stress-test components, or translate design files into functional UI. You can invoke specific skills to audit a pull request, generate color palettes, or render component states for testing.

Who it’s for

This is for design engineers and frontend developers who want to automate the repetitive aspects of UI maintenance and ensure their projects adhere to high standards of design and accessibility.

Highlights

  • Comprehensive UI reviews covering accessibility, layout, writing, typography, and color.
  • Automated accessibility checks against WCAG 2.2 standards.
  • Component stress-testing through automated state rendering and scenario testing.
  • Design-to-code capabilities that build UI from Figma files or images using existing tokens.
  • Tools for generating semantic color tokens and managing complex typography scales.
  • Support for reviewing active branches and pull requests to catch interface regressions.

Related

  • Project
    SWE-agent/mini-swe-agent

    A minimal AI software engineering agent that uses only bash commands to solve coding tasks, providing a high-performance, hackable alternative to complex agent frameworks.

  • Project
    pydantic/monty

    A secure, Rust-based Python sandbox designed for the high-speed execution of AI-generated code without the latency of containers.

  • Project
    MoonshotAI/Kimi-K3

    Kimi K3 is an open-weight, 2.8T-parameter native multimodal MoE model designed for long-horizon coding, deep research, and complex agentic knowledge work.

  • Project
    Nutlope/aicommits

    aicommits is a Node.js CLI that generates git commit messages with AI. It supports many providers, can reuse existing Claude/ChatGPT/OpenCode subscriptions without API keys, offers five message formats, custom prompts, multiple generated options, a Git hook, and full configuration via commands or environment variables. It only reads staged changes for privacy and safety.