claude-skills-collection
CLAUDE.md18 starsChanged 56 days agoMITView on GitHub
Read AGENTS.md for the shared repository rules before making changes. This file is the Claude Code entry point and carries the same skill inventory. This is a reusable skills collection for coding agents, focused on Swift and iOS development. It contains curated skills that provide specialized expertise across different iOS/Swift domains. Each skill is a self-contained directory with structured documentation and examples or reference material. The repository follows a consistent pattern: Each SKILL.md follows this frontmatter format: New skills use…
In this repository14 other files it gives its agents.
- AGENTS.mdAGENTS.md
- app-intents-widgetsSkill
- Cross-Platform App Development SkillSkill
- iOS Accessibility SkillSkill
- iOS Animation Graphics SkillSkill
- iphone-duo-designSkill
- Memory Leak Diagnosis SkillSkill
- swift-concurrency-migrationSkill
- swiftdata-core-data-migrationsSkill
- Swift Modern Architecture SkillSkill
- Swift Performance Optimization SkillSkill
- SpeechAnalyzer Framework ExpertSkill
- SwiftUI Programming SkillSkill
- Swift Unit Testing SkillSkill
The file
Show its rules and text
We split this CLAUDE.md into the rules it gives an agent
15 rules. About 0 are common, 10 are distinctive, and 5 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈10
Distinctive · estimate
This author's own choices. Worth a look.
5
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 3
Read AGENTS.md for the shared repository rules before making changes.
- Distinctive: line 58
When adding a new skill, follow these requirements:
- Very specific: line 63
Supporting Material: Add practical examples in examples/, or use docs/ and references/ when the skill is primarily a reference package
- Distinctive: line 65
Follow Swift naming conventions (camelCase, PascalCase for types)
- Distinctive: line 69
Use modern Swift features (async/await when appropriate)
- Very specific: line 70
Host Metadata: Add agents/openai.yaml with a concise display name, short description, and a default prompt that names $<skill-name>
- Very specific: line 71
Progressive Disclosure: Keep SKILL.md under 500 lines and link directly to focused examples or references
- Distinctive: line 77
All Swift code examples should adhere to:
- Distinctive: line 85
Explicit platform baselines: when APIs require particular Swift, Xcode, iOS, macOS, watchOS, or tvOS versions
- Distinctive: line 89
When skills include testing patterns:
- Distinctive: line 90
Prefer Swift Testing for new unit and integration tests when the target/toolchain supports it
- Very specific: line 102
Update SKILL.md and README.md with changes
- Distinctive: line 103
Test all code examples for correctness
- Very specific: line 118
When adding, removing, or renaming a skill, update the inventories or structural guidance in AGENTS.md, CLAUDE.md, CODEX.md, and README.md.
- Distinctive: line 122
Keep core skill procedures usable through ordinary file reads and shell tools.
≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
CLAUDE.md
Read AGENTS.md for the shared repository rules before making changes. This file is the Claude Code entry point and carries the same skill inventory.
Repository Overview
This is a reusable skills collection for coding agents, focused on Swift and iOS development. It contains curated skills that provide specialized expertise across different iOS/Swift domains. Each skill is a self-contained directory with structured documentation and examples or reference material.
Repository Structure
The repository follows a consistent pattern:
skills/<skill-name>/
├── SKILL.md # Skill definition with frontmatter and instructions
├── README.md # User-facing documentation
├── agents/
│ └── openai.yaml # Optional OpenAI host UI metadata
├── examples/ # Practical code and prompts (when appropriate)
│ ├── prompts.md # Positive and negative activation fixtures
│ └── *.swift # Working code examples
├── docs/ # Optional packaged documentation
└── references/ # Optional deep-dive guidesAvailable Skills
- swiftui-programming-skill - SwiftUI declarative UI development, SF Symbols, state management
- swift-modern-architecture-skill - Swift 6 architecture, SwiftData persistence, Observation framework, modern concurrency
- ios-accessibility-skill - VoiceOver, Dynamic Type, HIG compliance
- swift-performance-optimization-skill - Performance profiling, Instruments usage, memory efficiency
- cross-platform-app-development-skill - Multi-platform app strategies
- swift-unit-testing-skill - Swift Testing, XCTest, TDD, and test doubles
- ios-animation-graphics-skill - SwiftUI Canvas, Lottie animations
- memory-leak-diagnosis-skill - ARC, retain cycles, Instruments Leaks tool
- swift-SpeechAnalyzer-Framework-Expert - SpeechAnalyzer and SpeechTranscriber for on-device transcription on iOS 26+ and macOS 26+
- swift-concurrency-migration - Swift 6 concurrency migration and compiler-diagnostic workflows
- swiftdata-core-data-migrations - Safe SwiftData and Core Data schema/store migrations
- app-intents-widgets - App Intents, App Shortcuts, WidgetKit, and system surfaces
- iphone-duo-design - Adaptive iPhone Duo design, state continuity, scenes, camera, and readiness verification
Working with Skills
Skill File Structure
Each SKILL.md follows this frontmatter format:
---
name: <Skill Name>
description: <What the skill does, when to use it, and when not to use it>
---New skills use a lowercase kebab-case name matching their directory. Preserve the published name value of an existing skill unless a compatibility plan accompanies the change. Hosts use description for implicit activation, so front-load useful trigger terms and include exclusions.
Creating New Skills
When adding a new skill, follow these requirements:
- Frontmatter: Include only
nameand a scope-richdescription - Best Practices Section: Focused, domain-specific best practices
- Guidelines Section: Additional domain-specific guidelines with bullet points
- Supporting Material: Add practical examples in
examples/, or usedocs/andreferences/when the skill is primarily a reference package - Code Standards: Code examples must:
- Follow Swift naming conventions (camelCase, PascalCase for types)
- Include proper imports
- Be runnable/compilable where applicable
- Include explanatory comments where helpful
- Use modern Swift features (async/await when appropriate)
- Host Metadata: Add
agents/openai.yamlwith a concise display name, short description, and a default prompt that names$<skill-name> - Progressive Disclosure: Keep
SKILL.mdunder 500 lines and link directly to focused examples or references
New skill directories should use lowercase kebab-case. Existing published directories should not be renamed without a compatibility and migration plan.
Code Quality Standards
All Swift code examples should adhere to:
- Apple's Swift API Design Guidelines
- SwiftUI best practices for declarative UI code
- Modern concurrency patterns (async/await, actors) when applicable
- Memory safety patterns (weak/unowned references, avoiding retain cycles)
- Accessibility-first design (proper labels, hints, Dynamic Type support)
- Type safety and clear error handling
- Explicit platform baselines when APIs require particular Swift, Xcode, iOS, macOS, watchOS, or tvOS versions
Testing Examples
When skills include testing patterns:
- Prefer Swift Testing for new unit and integration tests when the target/toolchain supports it
- Retain XCTest for UI automation, performance tests, Objective-C tests, and legacy suites
- Follow Arrange-Act-Assert structure
- Include both sync and async test examples
- Show proper mocking/dependency injection patterns
- Demonstrate performance testing when relevant
Contributing
Reference CONTRIBUTING.md for contribution guidelines, which include:
- Fork and pull request workflow
- Code must follow Apple's Swift and iOS guidelines
- Update SKILL.md and README.md with changes
- Test all code examples for correctness
- For new skills, include complete directory structure
Development Notes
Validation, not an app build
This repository contains documentation and example code snippets, not a runnable app. There is no shared Package.swift or Xcode project; repository scripts validate structure and type-check standalone examples against their declared SDK targets.
Documentation-Focused
The primary artifacts are .md files with embedded Swift code blocks. Code examples are illustrative and designed to be copied into real projects.
Version Management
The plugin manifests carry the collection's semantic version. Individual portable skill frontmatter intentionally contains only name and description.
Repository Synchronization
When adding, removing, or renaming a skill, update the inventories or structural guidance in AGENTS.md, CLAUDE.md, CODEX.md, and README.md. Update plugin metadata when the change affects the plugin description or release version.
Host independence
Keep core skill procedures usable through ordinary file reads and shell tools. Put host-specific installation and invocation details in the compatibility guide. Use the installer for copy or symlink setup; never merge skill trees into existing unrelated directories.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

