Agent Skills: Testing Code

Testing Code

You are activating test generation and improvement capabilities. Your role is to ensure code has comprehensive, maintainable test coverage.

When to Activate

This skill activates when:

• New features or functions are implemented
• User requests tests ("add tests", "test this")
• Test coverage is low or missing
• Bugs are found (need regression tests)
• Refactoring code (need confidence tests pass)
• Before creating pull requests

Testing Philosophy

Why Test?

• Confidence: Know code works as expected
• Regression Prevention: Catch bugs from changes
• Documentation: Tests show how code should be used
• Design Feedback: Hard to test = bad design
• Refactoring Safety: Change with confidence

Test Pyramid

        /\
       /  \    E2E Tests (Few)
      /    \   - Test full workflows
     /------\
    /        \ Integration Tests (Some)
   /          \- Test module interactions
  /------------\
 /              \ Unit Tests (Many)
/______________/ - Test individual functions

Testing Process

1. Analyze Code

Understand what to test:

• What is the purpose?
• What are the inputs and outputs?
• What are the edge cases?
• What are the error conditions?
• What are the side effects?

2. Identify Test Cases

Happy Path

• Normal, expected usage
• Valid inputs producing valid outputs

Edge Cases

• Boundary values (0, -1, max, empty)
• Unusual but valid inputs
• Minimum and maximum values

Error Cases

• Invalid inputs
• Null/undefined/None
• Wrong types
• Violations of constraints

Integration Points

• External API calls
• Database interactions
• File system operations
• Other modules/classes

3. Write Tests

Follow this structure:

def test_function_name_condition_expected():
    """
    Test that function_name handles condition and returns expected result.
    """
    # Arrange: Set up test data and conditions
    input_data = ...
    expected_output = ...

    # Act: Execute the code under test
    result = function_name(input_data)

    # Assert: Verify the result
    assert result == expected_output

4. Check Coverage

Ensure comprehensive coverage:

• All code paths executed
• All branches tested
• All error conditions tested
• Edge cases covered

Test Patterns

Unit Tests

Test individual functions in isolation.

Python Example:

import pytest
from mymodule import calculate_total

def test_calculate_total_with_valid_items():
    """Calculate total for valid items."""
    items = [
        {"price": 10.00, "quantity": 2},
        {"price": 5.00, "quantity": 3},
    ]
    assert calculate_total(items) == 35.00

def test_calculate_total_with_empty_list():
    """Calculate total returns 0 for empty list."""
    assert calculate_total([]) == 0.00

def test_calculate_total_with_zero_quantity():
    """Calculate total handles zero quantity."""
    items = [{"price": 10.00, "quantity": 0}]
    assert calculate_total(items) == 0.00

def test_calculate_total_raises_on_negative_price():
    """Calculate total raises ValueError for negative price."""
    items = [{"price": -10.00, "quantity": 1}]
    with pytest.raises(ValueError, match="Price cannot be negative"):
        calculate_total(items)

JavaScript Example:

import { describe, it, expect } from "vitest";
import { calculateTotal } from "./cart";

describe("calculateTotal", () => {
  it("calculates total for valid items", () => {
    const items = [
      { price: 10.0, quantity: 2 },
      { price: 5.0, quantity: 3 },
    ];
    expect(calculateTotal(items)).toBe(35.0);
  });

  it("returns 0 for empty list", () => {
    expect(calculateTotal([])).toBe(0);
  });

  it("handles zero quantity", () => {
    const items = [{ price: 10.0, quantity: 0 }];
    expect(calculateTotal(items)).toBe(0);
  });

  it("throws error for negative price", () => {
    const items = [{ price: -10.0, quantity: 1 }];
    expect(() => calculateTotal(items)).toThrow("Price cannot be negative");
  });
});

Integration Tests

Test multiple components working together.

Example:

def test_user_registration_flow():
    """Test complete user registration workflow."""
    # Arrange: Set up test database
    db = setup_test_database()
    api = UserAPI(db)

    # Act: Register user
    response = api.register({
        "email": "test@example.com",
        "password": "SecurePass123!",
    })

    # Assert: User created and email sent
    assert response.status == 201
    assert db.users.find_by_email("test@example.com") is not None
    assert email_was_sent("test@example.com", "Welcome")

Mocking & Stubbing

Isolate code from external dependencies.

Python Example:

from unittest.mock import Mock, patch

def test_send_notification_calls_email_service():
    """Test that send_notification calls email service correctly."""
    # Arrange: Mock email service
    mock_email = Mock()

    # Act: Send notification
    with patch('mymodule.email_service', mock_email):
        send_notification("user@example.com", "Hello")

    # Assert: Email service called with correct args
    mock_email.send.assert_called_once_with(
        to="user@example.com",
        subject="Notification",
        body="Hello"
    )

Parameterized Tests

Test multiple inputs efficiently.

Python Example:

@pytest.mark.parametrize("input,expected", [
    (0, "zero"),
    (1, "one"),
    (2, "two"),
    (10, "many"),
    (-5, "negative"),
])
def test_number_to_word(input, expected):
    """Test number_to_word for various inputs."""
    assert number_to_word(input) == expected

Property-Based Tests

Generate test cases automatically.

Python Example:

from hypothesis import given
from hypothesis.strategies import integers

@given(integers())
def test_double_is_reversible(x):
    """Test that doubling and halving returns original value."""
    assert half(double(x)) == x

Test Organization

Directory Structure

project/
├── src/
│   └── mymodule/
│       ├── __init__.py
│       ├── calculator.py
│       └── user.py
└── tests/
    ├── __init__.py
    ├── unit/
    │   ├── test_calculator.py
    │   └── test_user.py
    ├── integration/
    │   └── test_user_flow.py
    └── conftest.py  # Shared fixtures

Naming Conventions

• Test files: test_*.py or *_test.py
• Test functions: test_<function>_<condition>_<expected>
• Test classes: Test<ClassName>

Good Names:

• test_calculate_total_with_empty_list_returns_zero
• test_user_login_with_invalid_password_raises_error
• test_api_returns_404_for_nonexistent_resource

Bad Names:

• test_calc (too vague)
• test_1 (no meaning)
• test_it_works (what works?)

Testing Checklist

For each function/class, ensure tests for:

Functionality

• [ ] Happy path (normal usage)
• [ ] Return values correct
• [ ] Side effects occur as expected

Edge Cases

• [ ] Empty inputs ([], "", 0, None)
• [ ] Boundary values (min, max, -1)
• [ ] Large inputs
• [ ] Special characters/values

Error Handling

• [ ] Invalid inputs
• [ ] Null/undefined/None
• [ ] Wrong types
• [ ] Constraint violations
• [ ] External failures (API down, etc.)

Integration

• [ ] Calls to other functions/modules
• [ ] Database interactions
• [ ] File operations
• [ ] Network requests

Test Quality Standards

Good Tests Are:

Fast:

• Run in milliseconds
• No unnecessary I/O
• Use mocks for external dependencies

Independent:

• No shared state between tests
• Can run in any order
• Don't depend on other tests

Repeatable:

• Same result every time
• No flaky tests
• No dependency on external state

Self-Validating:

• Pass or fail clearly
• No manual inspection
• Descriptive failure messages

Timely:

• Written before or with code (TDD)
• Not as an afterthought
• Updated when code changes

Anti-Patterns to Avoid

Testing Implementation Details:

# Bad: Tests internal state
def test_cache_uses_dict():
    cache = Cache()
    assert isinstance(cache._storage, dict)

# Good: Tests behavior
def test_cache_stores_and_retrieves_value():
    cache = Cache()
    cache.set("key", "value")
    assert cache.get("key") == "value"

Overly Complex Tests:

# Bad: Test is hard to understand
def test_complex():
    data = [process(x) for x in range(100) if x % 2]
    result = calculate(data)
    assert result == sum(data) / len(data)

# Good: Clear and simple
def test_calculate_returns_average():
    data = [2, 4, 6]
    assert calculate(data) == 4.0

Not Testing Edge Cases:

# Incomplete: Only tests happy path
def test_divide():
    assert divide(10, 2) == 5

# Complete: Tests edge cases too
def test_divide_with_valid_numbers():
    assert divide(10, 2) == 5

def test_divide_by_zero_raises_error():
    with pytest.raises(ZeroDivisionError):
        divide(10, 0)

def test_divide_with_negative_numbers():
    assert divide(-10, 2) == -5

Coverage Targets

Minimum Coverage

• Critical paths: 100%
• Business logic: 95%+
• Utilities: 90%+
• UI code: 70%+

Focus Areas

1. Business-critical features
2. Complex algorithms
3. Error handling
4. Security-sensitive code
5. Bug-prone areas

Don't Obsess Over 100%

• Some code is hard to test (UI rendering, etc.)
• Focus on meaningful coverage
• Quality > quantity

Integration Points

Invokes

• Tester Agent: Core testing logic from ~/.amplihack/.claude/agents/tester.md
• Code Analysis: To identify untested code
• Coverage Tools: pytest-cov, coverage.py, jest --coverage

Escalates To

• Builder Agent: To fix code that's hard to test
• Reviewer Agent: To review test quality

References

• Testing Guides: Language-specific best practices
• Test Patterns: Common testing patterns
• Fixtures: Shared test data and setup

Example Workflow

User: "I just implemented this authentication function. Can you add tests?"

Testing Code Skill:

Let me analyze the authentication function and create comprehensive tests.

## Function Analysis

The `authenticate(username, password)` function:
- Returns JWT token on success
- Raises `InvalidCredentialsError` on bad password
- Raises `UserNotFoundError` if user doesn't exist
- Hashes password before comparison
- Logs failed attempts

## Test Plan

I'll create tests for:
1. Successful authentication (happy path)
2. Invalid password (error case)
3. Non-existent user (error case)
4. Empty username/password (edge case)
5. SQL injection attempt (security)
6. Logging of failed attempts (side effect)

## Generated Tests

```python
import pytest
from unittest.mock import Mock, patch
from myapp.auth import authenticate, InvalidCredentialsError, UserNotFoundError

class TestAuthenticate:
    """Tests for authenticate function."""

    def test_successful_authentication_returns_jwt_token(self, mock_db):
        """Authenticate returns valid JWT for correct credentials."""
        # Arrange
        mock_db.get_user.return_value = User(
            username="alice",
            password_hash="$2b$12$..."
        )

        # Act
        token = authenticate("alice", "correct_password")

        # Assert
        assert token is not None
        assert jwt.decode(token)["username"] == "alice"

    def test_invalid_password_raises_error(self, mock_db):
        """Authenticate raises InvalidCredentialsError for wrong password."""
        mock_db.get_user.return_value = User(
            username="alice",
            password_hash="$2b$12$..."
        )

        with pytest.raises(InvalidCredentialsError):
            authenticate("alice", "wrong_password")

    def test_nonexistent_user_raises_error(self, mock_db):
        """Authenticate raises UserNotFoundError for unknown user."""
        mock_db.get_user.return_value = None

        with pytest.raises(UserNotFoundError):
            authenticate("nobody", "password")

    def test_empty_username_raises_error(self):
        """Authenticate raises ValueError for empty username."""
        with pytest.raises(ValueError, match="Username cannot be empty"):
            authenticate("", "password")

    def test_logs_failed_authentication_attempts(self, mock_db, mock_logger):
        """Failed authentication attempts are logged."""
        mock_db.get_user.return_value = User(username="alice", ...)

        try:
            authenticate("alice", "wrong_password")
        except InvalidCredentialsError:
            pass

        mock_logger.warning.assert_called_with(
            "Failed login attempt for user: alice"
        )

@pytest.fixture
def mock_db():
    """Mock database for tests."""
    return Mock()

@pytest.fixture
def mock_logger():
    """Mock logger for tests."""
    with patch('myapp.auth.logger') as mock:
        yield mock

Coverage Report

Running tests:

pytest tests/test_auth.py --cov=myapp.auth --cov-report=term-missing

Current coverage: 95% (missing: edge case for special characters)

Suggestions

Consider adding:

1. Test for rate limiting (if implemented)
2. Test for account lockout after X failures
3. Performance test for password hashing time

Ready to implement these tests?

## Success Criteria

Good test suite:
- All critical paths covered
- Edge cases tested
- Tests are fast (<1s per file)
- Tests are independent
- Clear test names
- Failures are descriptive
- Easy to add new tests

## Related Capabilities

- **Agent**: Tester agent (this skill invokes it)
- **Skill**: "Reviewing Code" for test quality review
- **Skill**: "Debugging Issues" when tests fail
- **Documentation**: Language-specific testing guides

---

Remember: Tests are documentation that proves your code works. Write tests you'll thank yourself for later.