Expert guidance for any AI coding tool that supports the Agent Skills open format — Swift Testing framework, test doubles, fixtures, and testing best practices.
- Teams adopting Swift Testing framework who need modern testing patterns.
- Developers writing unit tests, integration tests, or snapshot tests.
- Anyone wanting reliable test infrastructure (mocks, stubs, spies, fixtures).
- Teams migrating from XCTest to Swift Testing.
Install this skill with a single command:
npx skills add https://github.com/bocato/swift-testing-agent-skill --skill swift-testingThen use the skill in your AI agent, for example:
Use the swift testing skill and help me write tests for the UserService class
GitHub Copilot supports Agent Skills in VS Code, VS Code Insiders, and Copilot CLI.
Copy the swift-testing/ folder to your repository's .github/skills/ directory:
# Clone this repo and copy the skill
git clone https://github.com/bocato/swift-testing-agent-skill.git
cp -r swift-testing-agent-skill/swift-testing .github/skills/Your project structure should look like:
your-project/
├── .github/
│ └── skills/
│ └── swift-testing/
│ ├── SKILL.md
│ └── references/
│ └── ...
└── ...
For personal use across all projects, copy to your home directory:
# Create personal skills directory if it doesn't exist
mkdir -p ~/.copilot/skills
# Copy the skill
cp -r swift-testing-agent-skill/swift-testing ~/.copilot/skills/Ensure Agent Skills are enabled in VS Code settings:
{
"chat.useAgentSkills": true
}Copilot will automatically discover and load the skill when relevant to your testing tasks.
To install this Skill for your personal use in Claude Code:
-
Add the marketplace:
/plugin marketplace add bocato/swift-testing-agent-skill
-
Install the Skill:
/plugin install swift-testing@swift-testing-agent-skill
To automatically provide this Skill to everyone working in a repository, configure the repository's .claude/settings.json:
{
"enabledPlugins": {
"swift-testing@swift-testing-agent-skill": true
},
"extraKnownMarketplaces": {
"swift-testing-agent-skill": {
"source": {
"source": "github",
"repo": "bocato/swift-testing-agent-skill"
}
}
}
}When team members open the project, Claude Code will prompt them to install the Skill.
Gemini CLI supports custom instructions through GEMINI.md files or the .gemini/ directory.
Copy the skill content to your project's .gemini/ directory:
# Clone this repo
git clone https://github.com/bocato/swift-testing-agent-skill.git
# Create .gemini directory and copy the skill
mkdir -p .gemini/skills
cp -r swift-testing-agent-skill/swift-testing .gemini/skills/Your project structure should look like:
your-project/
├── .gemini/
│ └── skills/
│ └── swift-testing/
│ ├── SKILL.md
│ └── references/
│ └── ...
└── ...
For personal use across all projects, copy to your home directory:
# Create personal Gemini skills directory
mkdir -p ~/.gemini/skills
# Copy the skill
cp -r swift-testing-agent-skill/swift-testing ~/.gemini/skills/Gemini CLI will automatically discover skills in these locations when relevant to your testing tasks.
- Clone this repository.
- Install or symlink the
swift-testing/folder following your tool's official skills installation docs (see links below). - Use your AI tool as usual and ask it to use the "swift-testing" skill for testing tasks.
Follow your tool's official documentation, here are a few popular ones:
- Gemini CLI: Use
.gemini/skills/(project) or~/.gemini/skills/(personal) - GitHub Copilot: About Agent Skills - Use
.github/skills/(project) or~/.copilot/skills/(personal) - Codex: Where to save skills
- Claude: Using Skills
- Cursor: Enabling Skills
How to verify:
Your agent should reference the triage/playbook in swift-testing/SKILL.md and jump into the relevant reference file for your testing question or task.
This skill gives your AI coding tool comprehensive Swift Testing guidance. It can:
- Choose the right assertion (
#expectvs#require) - Structure tests with Arrange-Act-Assert pattern
- Apply F.I.R.S.T. principles (Fast, Isolated, Repeatable, Self-Validating, Timely)
- Follow the test pyramid (80% unit, 15% integration, 5% UI)
- Understand Martin Fowler's test double taxonomy
- Create Dummies, Fakes, Stubs, Spies, SpyingStubs, and Mocks
- Choose between state verification and behavior verification
- Place test doubles close to interfaces with
#if DEBUG
- Build fixtures with sensible defaults
- Place fixtures close to models
- Handle dates deterministically
- Organize test suites with tags and traits
- Write parameterized tests for multiple inputs
- Test async code with proper patterns
- Use confirmation for callback/delegate testing
- Set up snapshot testing for UI regression
- Complete XCTest to Swift Testing migration guide
- Assertion mapping (XCTAssertEqual -> #expect)
- Setup/teardown patterns (setUp -> init)
- Async test migration patterns
Practical Focus: Based on real-world testing patterns from production codebases, focusing on what developers actually need.
Proper Taxonomy: Uses Martin Fowler's official test double terminology, clarifying the confusion between "mocks" and "spies" in the Swift community.
Non-Opinionated: Focuses on industry-standard best practices like F.I.R.S.T. principles and Arrange-Act-Assert, not architectural preferences.
Swift Testing Ready: Covers the modern Swift Testing framework with @Test, #expect, #require, and @Suite.
Infrastructure Guidance: Includes patterns for test doubles, fixtures, and their proper placement in your codebase.
swift-testing/
├── SKILL.md # Main skill file with decision trees
└── references/
├── _index.md # Quick navigation for all topics
├── test-organization.md # Suites, tags, traits, parallel execution
├── parameterized-tests.md # Testing multiple inputs efficiently
├── async-testing.md # Async patterns, confirmation, timeouts
├── migration-xctest.md # XCTest to Swift Testing migration
├── test-doubles.md # Dummy, Fake, Stub, Spy, SpyingStub, Mock
├── fixtures.md # Fixture patterns and placement
├── integration-testing.md # Module interaction testing
└── snapshot-testing.md # UI regression testing
import Testing
@Test("User can be created with valid data")
func createUser() {
let user = User(name: "Alice", age: 30)
#expect(user.name == "Alice")
#expect(user.age == 30)
}@Test func calculateTotal() {
// Given
let cart = ShoppingCart()
cart.add(Item(price: 10))
// When
let total = cart.calculateTotal()
// Then
#expect(total == 10)
}#if DEBUG
final class UserRepositorySpyingStub: UserRepositoryProtocol {
// Spy
private(set) var savedUsers: [User] = []
// Stub
var usersToReturn: [User] = []
func save(_ user: User) async throws {
savedUsers.append(user)
}
func getAll() async throws -> [User] {
usersToReturn
}
}
#endif#if DEBUG
extension User {
static func fixture(
id: UUID = UUID(),
name: String = "Test User",
email: String = "[email protected]"
) -> User {
User(id: id, name: name, email: email)
}
}
#endifContributions are welcome! This repository follows the Agent Skills open format, which has specific structural requirements.
We strongly recommend using AI assistance for contributions:
- Use the skill-creator skill with Claude to ensure proper formatting
- This helps maintain the Agent Skills format and ensures your contribution works correctly with AI agents
Please read CONTRIBUTING.md for:
- How to use the skill-creator skill for contributions
- Agent Skills format requirements
- Quality standards and best practices
- Pull request process
This skill is maintained to reflect the latest Swift Testing best practices and will be updated as the framework evolves.
This skill is open-source and available under the MIT License. See LICENSE for details.