Summary

This project provides an "agent skill" designed to enhance AI coding assistants like Claude Code, Codex, Gemini, and Cursor for SwiftUI development. It aims to help developers write smarter, simpler, and more modern SwiftUI code by offering guidance on API usage, design, performance, and accessibility, directly addressing common errors made by AI models.

How It Works

The skill leverages the "Agent Skills format" to integrate practical knowledge and experience into AI coding tools. This approach allows AI agents to access specific rules and best practices, enabling them to generate more accurate and efficient SwiftUI code by avoiding pitfalls such as accessibility issues, deprecated API usage, and performance regressions.

Quick Start & Requirements

• Installation: Use npx skills add https://github.com/twostraws/swiftui-agent-skill --skill swiftui-pro.
• Prerequisites: Requires Node.js. If npx is not found, install Node.js via Homebrew (brew install node).
• Usage: Trigger the skill using commands like /swiftui-pro (Claude Code) or $swiftui-pro (Codex), or via natural language prompts.
• Configuration: Can be installed for a single project or globally. Alternatively, the repository can be cloned for custom installation.
• Documentation: Further details available in the creator's article: SwiftUI Agent Skill - Write better code with Claude, Codex, and other AI tools.

Highlighted Details

• Targets common LLM mistakes, including invisible VoiceOver elements, deprecated API usage, and performance problems.
• Built upon thousands of hours of real-world SwiftUI development experience.
• Compatible with multiple AI coding platforms: Claude Code, Codex, Gemini, Cursor.

Maintenance & Community

Created by Paul Hudson, known for Hacking with Swift. Contributions are welcomed for new checks, improvements, and documentation. The project emphasizes concise contributions to respect user token budgets.

Licensing & Compatibility

Licensed under the MIT License, which permits commercial use, modification, distribution, and private use.

Limitations & Caveats

Using agent skills incurs a token cost, particularly for extensive checks. The skill focuses on edge cases, surprising behaviors, and soft deprecations rather than fundamental LLM knowledge.