TDD with Vitest and TypeScript

Guide Claude through Test-Driven Development workflows using Vitest and TypeScript.

Core TDD Cycle: Red-Green-Refactor

Always follow this three-phase cycle:

1. Red: Write a failing test that defines desired behavior
2. Green: Write minimal code to make the test pass
3. Refactor: Improve code quality while keeping tests green

Workflow Pattern

// 1. RED: Write the test first
describe('Calculator', () => {
  it('adds two numbers', () => {
    const calc = new Calculator();
    expect(calc.add(2, 3)).toBe(5);
  });
});

// Run test → Watch it fail (Red)
// 2. GREEN: Implement minimal code
class Calculator {
  add(a: number, b: number): number {
    return a + b;
  }
}

// Run test → Watch it pass (Green)
// 3. REFACTOR: Improve if needed while keeping tests green

Vitest Setup and Configuration

Basic Vitest Config

Create vitest.config.ts:

import { defineConfig } from 'vitest/config';

export default defineConfig({
  test: {
    globals: true,
    environment: 'node', // or 'jsdom' for DOM testing
    coverage: {
      provider: 'v8',
      reporter: ['text', 'json', 'html'],
    },
  },
});

TypeScript Configuration

Ensure tsconfig.json includes:

{
  "compilerOptions": {
    "types": ["vitest/globals"],
    "esModuleInterop": true,
    "skipLibCheck": true
  }
}

Test File Organization

Naming Conventions

• Test files: *.test.ts or *.spec.ts
• Place tests adjacent to source files or in __tests__ directories
• Match test file names to source files: calculator.ts → calculator.test.ts

Structure Pattern

import { describe, it, expect, beforeEach, afterEach } from 'vitest';

describe('FeatureName', () => {
  // Setup
  beforeEach(() => {
    // Runs before each test
  });

  afterEach(() => {
    // Cleanup after each test
  });

  describe('specific behavior', () => {
    it('does something specific', () => {
      // Arrange
      const input = setupTestData();
      
      // Act
      const result = performAction(input);
      
      // Assert
      expect(result).toBe(expected);
    });
  });
});

TypeScript Testing Patterns

Type-Safe Test Data

interface User {
  id: number;
  name: string;
  email: string;
}

function createTestUser(overrides?: Partial<User>): User {
  return {
    id: 1,
    name: 'Test User',
    email: 'test@example.com',
    ...overrides,
  };
}

it('processes user data', () => {
  const user = createTestUser({ name: 'Custom Name' });
  expect(processUser(user)).toBeDefined();
});

Testing Generic Functions

describe('generic array utilities', () => {
  it('filters array by predicate', () => {
    const numbers = [1, 2, 3, 4, 5];
    const result = filter(numbers, (n: number) => n > 3);
    expect(result).toEqual([4, 5]);
  });
});

Testing Async/Promise Code

describe('async operations', () => {
  it('fetches user data', async () => {
    const user = await fetchUser(1);
    expect(user.id).toBe(1);
  });

  it('handles errors', async () => {
    await expect(fetchUser(-1)).rejects.toThrow('Invalid ID');
  });
});

Mocking and Stubbing

Module Mocks

import { vi } from 'vitest';
import { fetchData } from './api';

vi.mock('./api');

describe('data processing', () => {
  it('processes fetched data', async () => {
    vi.mocked(fetchData).mockResolvedValue({ data: 'test' });
    
    const result = await processData();
    expect(result).toBe('processed: test');
  });
});

Function Spies

describe('event handling', () => {
  it('calls callback on event', () => {
    const callback = vi.fn();
    const handler = new EventHandler(callback);
    
    handler.trigger('test-event');
    
    expect(callback).toHaveBeenCalledWith('test-event');
    expect(callback).toHaveBeenCalledTimes(1);
  });
});

Partial Mocks

import * as utils from './utils';

vi.spyOn(utils, 'helperFunction').mockReturnValue('mocked');

it('uses mocked helper', () => {
  const result = mainFunction();
  expect(utils.helperFunction).toHaveBeenCalled();
  expect(result).toContain('mocked');
});

Common Testing Patterns

Testing Classes

describe('UserService', () => {
  let service: UserService;
  let mockRepository: MockRepository;

  beforeEach(() => {
    mockRepository = new MockRepository();
    service = new UserService(mockRepository);
  });

  it('creates user with valid data', async () => {
    const userData = { name: 'John', email: 'john@example.com' };
    
    const user = await service.createUser(userData);
    
    expect(user.id).toBeDefined();
    expect(mockRepository.save).toHaveBeenCalledWith(
      expect.objectContaining(userData)
    );
  });
});

Testing Pure Functions

describe('pure utility functions', () => {
  it('capitalizes first letter', () => {
    expect(capitalize('hello')).toBe('Hello');
    expect(capitalize('')).toBe('');
    expect(capitalize('WORLD')).toBe('WORLD');
  });
});

Testing Error Handling

describe('error scenarios', () => {
  it('throws on invalid input', () => {
    expect(() => divide(10, 0)).toThrow('Division by zero');
  });

  it('returns error result', () => {
    const result = parseJSON('invalid json');
    expect(result.success).toBe(false);
    expect(result.error).toBeDefined();
  });
});

Parametric Tests

import { describe, it, expect } from 'vitest';

describe.each([
  { input: 2, expected: 4 },
  { input: 3, expected: 9 },
  { input: 4, expected: 16 },
])('square function', ({ input, expected }) => {
  it(`squares ${input} to ${expected}`, () => {
    expect(square(input)).toBe(expected);
  });
});

Test Coverage Guidelines

Running Coverage

vitest --coverage

Coverage Targets

• Aim for 80%+ coverage on business logic
• 100% coverage on critical paths (authentication, payments, etc.)
• Don't obsess over 100% everywhere—focus on meaningful tests

What to Test

Always test:

• Business logic and domain rules
• Error handling and edge cases
• Public APIs and interfaces
• Data transformations

Consider skipping:

• Simple getters/setters
• Framework/library code
• Trivial type definitions
• Configuration files

TDD Best Practices

Write Tests First

Always start with the test, not the implementation:

// ❌ BAD: Writing implementation first
class Calculator {
  add(a: number, b: number) { return a + b; }
}

// ✅ GOOD: Test first
it('adds two numbers', () => {
  expect(new Calculator().add(2, 3)).toBe(5);
});

One Assertion Per Test

Keep tests focused:

// ❌ BAD: Multiple concerns
it('user operations', () => {
  const user = createUser();
  expect(user.id).toBeDefined();
  expect(updateUser(user)).toBeTruthy();
  expect(deleteUser(user.id)).toBeUndefined();
});

// ✅ GOOD: Single concern
it('creates user with ID', () => {
  const user = createUser();
  expect(user.id).toBeDefined();
});

it('updates existing user', () => {
  const user = createUser();
  expect(updateUser(user)).toBeTruthy();
});

Test Behavior, Not Implementation

// ❌ BAD: Testing implementation details
it('calls internal helper method', () => {
  const service = new Service();
  const spy = vi.spyOn(service as any, '_internalHelper');
  service.process();
  expect(spy).toHaveBeenCalled();
});

// ✅ GOOD: Testing behavior
it('processes data correctly', () => {
  const service = new Service();
  const result = service.process(inputData);
  expect(result).toEqual(expectedOutput);
});

Keep Tests Fast

• Use mocks for external dependencies (databases, APIs, file system)
• Avoid sleep/setTimeout in tests
• Run expensive setup once with beforeAll when safe

Descriptive Test Names

// ❌ BAD: Vague
it('works', () => { /* ... */ });

// ✅ GOOD: Descriptive
it('returns empty array when no users match filter criteria', () => {
  /* ... */
});

Common TDD Workflow

Starting a New Feature

1. Write a high-level test describing the feature:

describe('User Registration', () => {
  it('creates new user account with valid email', async () => {
    const result = await registerUser({
      email: 'new@example.com',
      password: 'secure123',
    });
    
    expect(result.success).toBe(true);
    expect(result.user.email).toBe('new@example.com');
  });
});

2. Run test → See it fail (Red)
3. Implement minimal code → See it pass (Green)
4. Add edge case tests:

it('rejects registration with existing email', async () => {
  await registerUser({ email: 'existing@example.com', password: 'pass' });
  
  const result = await registerUser({
    email: 'existing@example.com',
    password: 'pass2',
  });
  
  expect(result.success).toBe(false);
  expect(result.error).toContain('Email already registered');
});

it('rejects weak passwords', async () => {
  const result = await registerUser({
    email: 'new@example.com',
    password: '123',
  });
  
  expect(result.success).toBe(false);
  expect(result.error).toContain('Password too weak');
});

5. Refactor implementation while keeping tests green

Debugging Failed Tests

When tests fail unexpectedly:

1. Check test isolation—are tests interfering with each other?
2. Verify mocks are properly reset between tests
3. Use it.only() to run single test
4. Add console.log or debugger statements
5. Check async timing issues

Integration Testing with Vitest

Testing Multiple Units Together

describe('Order Processing Integration', () => {
  let database: TestDatabase;
  let paymentGateway: MockPaymentGateway;
  let orderService: OrderService;

  beforeEach(async () => {
    database = await TestDatabase.create();
    paymentGateway = new MockPaymentGateway();
    orderService = new OrderService(database, paymentGateway);
  });

  afterEach(async () => {
    await database.cleanup();
  });

  it('completes order flow from cart to confirmation', async () => {
    const user = await database.createUser();
    const cart = await orderService.createCart(user.id);
    await orderService.addItem(cart.id, { productId: 1, quantity: 2 });
    
    paymentGateway.simulateSuccess();
    const order = await orderService.checkout(cart.id);
    
    expect(order.status).toBe('confirmed');
    expect(order.items).toHaveLength(1);
  });
});

Watch Mode

Vitest runs in watch mode by default during development:

vitest

This automatically re-runs tests when files change, enabling rapid TDD cycles.

Quick Reference: Common Matchers

// Equality
expect(value).toBe(5);                    // Strict equality (===)
expect(value).toEqual({ a: 1 });          // Deep equality

// Truthiness
expect(value).toBeTruthy();
expect(value).toBeFalsy();
expect(value).toBeNull();
expect(value).toBeUndefined();
expect(value).toBeDefined();

// Numbers
expect(value).toBeGreaterThan(3);
expect(value).toBeLessThan(10);
expect(value).toBeCloseTo(0.3);           // Floating point

// Strings
expect(string).toMatch(/pattern/);
expect(string).toContain('substring');

// Arrays
expect(array).toContain(item);
expect(array).toHaveLength(3);

// Objects
expect(object).toHaveProperty('key');
expect(object).toMatchObject({ a: 1 });   // Partial match

// Exceptions
expect(() => fn()).toThrow();
expect(() => fn()).toThrow('Error message');
expect(async () => fn()).rejects.toThrow();

// Functions
expect(fn).toHaveBeenCalled();
expect(fn).toHaveBeenCalledWith(arg1, arg2);
expect(fn).toHaveBeenCalledTimes(2);

Tips for Effective TDD

1. Start simple: Begin with the simplest test case, not the most complex
2. Take small steps: Write one test, make it pass, refactor, repeat
3. Trust the process: Resist urge to write implementation before tests
4. Refactor fearlessly: With good test coverage, refactoring is safe
5. Keep tests maintainable: Tests are code too—keep them clean and DRY
6. Run tests frequently: Vitest's watch mode makes this effortless
7. Write tests for bugs: When you find a bug, write a test that exposes it first

When to Use This Skill

Apply TDD when:

• Building new features from scratch
• Fixing bugs (write failing test first)
• Refactoring existing code
• Learning a new API or library
• Working on critical business logic

TDD is especially valuable for:

• Pure functions and algorithms
• Business logic and domain models
• Data transformations
• API endpoints and services