swift-patterns-skill

Swift Patterns Skill

A comprehensive Swift/SwiftUI knowledge base for AI coding tools, following the Agent Skills standard.

Provides expert guidance on state management, navigation, view composition, performance optimization, and modern SwiftUI API usage to help AI assistants generate better SwiftUI code.

What This Skill Provides

Comprehensive SwiftUI expertise across:

Core Topics

• State Management – Property wrapper selection (@State, @Binding, @Observable), ownership rules, data flow patterns
• Modern APIs – iOS 17/18/26 replacements for deprecated APIs, complete migration guides
• View Composition – Extraction patterns, parent/child data flow, view identity and performance
• Navigation – NavigationStack, sheets, deep linking, type-safe routing patterns

Advanced Areas

• Lists & Collections – Stable identity with ForEach, pagination, lazy containers
• Performance Optimization – View optimization strategies, avoiding recomputation, memory management
• Testing & Dependency Injection – Protocol-based patterns, test doubles, testable architecture
• Code Quality – Refactoring playbooks, code smell detection, anti-pattern identification

All guidance is based on Apple's official documentation and focuses on facts over opinions – no architectural mandates.

Quick Start

Install with a single command:

npx skills add https://github.com/efremidze/swift-patterns-skill --skill swift-patterns

Then use it in your AI assistant:

Review my SwiftUI view for state management issues

View on skills.sh →

Installation

Recommended: Using skills.sh CLI

The easiest way to install:

npx skills add https://github.com/efremidze/swift-patterns-skill --skill swift-patterns

This installs the skill and makes it available to your AI assistant.

Alternative: Claude Code Plugin

For Claude Code users, add via the marketplace:

1. Add the marketplace:

   /plugin marketplace add efremidze/swift-patterns-skill
   

2. Install the skill:

   /plugin install swift-patterns@swift-patterns-skill
   

Or configure for your team in .claude/settings.json:

{
  "enabledPlugins": {
    "swift-patterns@swift-patterns-skill": true
  },
  "extraKnownMarketplaces": {
    "swift-patterns-skill": {
      "source": {
        "source": "github",
        "repo": "efremidze/swift-patterns-skill"
      }
    }
  }
}

Manual Installation

If you prefer manual setup:

1. Clone this repository
2. Install or symlink swift-patterns/ to your tool's skills directory
3. Configure your AI tool to use swift-patterns

Skill Structure

The skill follows a progressive disclosure model—core workflows in SKILL.md, detailed guidance in references/:

swift-patterns/
  SKILL.md                          # Entry point: workflow routing, quick refs, review checklist
  references/
    state.md                        # Property wrappers, ownership, @Observable patterns
    navigation.md                   # NavigationStack, sheets, deep linking
    view-composition.md             # View extraction, data flow patterns
    lists-collections.md            # ForEach identity, List vs LazyVStack
    scrolling.md                    # Pagination, scroll position management
    concurrency.md                  # .task modifier, async lifecycle
    performance.md                  # View optimization, lazy loading strategies
    testing-di.md                   # Dependency injection, test doubles
    patterns.md                     # Container views, ViewModifiers, PreferenceKeys
    modern-swiftui-apis.md          # iOS 17/18/26 API replacements and migration
    refactor-playbooks.md           # Step-by-step refactoring guides
    workflows-review.md             # Review methodology and standards
    workflows-refactor.md           # Refactoring methodology, invariants
    code-review-refactoring.md      # Code smells, anti-patterns, quality checks

Related Projects

Other Skills

• swift-architecture-skill – Architectural patterns and project structure guidance (complements this skill's focus on SwiftUI patterns)

Dynamic Runtime Tools

• swift-patterns-mcp – MCP server with intelligent search, retrieval, and persistent memory