Playwright E2E Tester

Overview

Expert in end-to-end testing with Playwright, the modern cross-browser testing framework. Specializes in test generation, page object patterns, visual regression testing, and CI/CD integration. Handles complex testing scenarios including authentication flows, API mocking, and mobile emulation.

When to Use

• Setting up Playwright in a new or existing project
• Writing E2E tests for critical user flows
• Debugging flaky tests or test failures
• Implementing visual regression testing
• Configuring Playwright for CI/CD pipelines
• Migrating from Cypress, Selenium, or Puppeteer

Decision Points

Locator Strategy Selection

Is element semantic (button, heading, form control)?
├─ YES → Use role-based locators (getByRole, getByLabel)
│   └─ Expected to change frequently?
│       ├─ NO → Stop here (most stable)
│       └─ YES → Add getByTestId as backup
└─ NO → Is element purely presentational?
    ├─ YES → Use getByTestId (add data-testid attribute)
    │   └─ Can't modify markup?
    │       └─ Use CSS selector (last resort, document fragility)
    └─ NO → Use getByText for content-based selection

Wait Strategy Decision

Action type:
├─ Navigation (page.goto, click link) → Use waitForURL()
├─ Element appears/disappears → Use waitForSelector() or visibility assertions
├─ API response affects UI → Use waitForResponse() then element assertion
├─ Animation/transition → Use waitForFunction() with custom condition
└─ Network request completion → Use page.route() with route.fulfill()

Test Organization

Test complexity:
├─ Single page interaction → Inline test with direct selectors
├─ Multi-step flow → Use Page Object Model pattern
├─ Cross-page workflow → Use fixtures for shared state
└─ Multi-app integration → Use projects with different configs

Failure Modes

Stale Element Reference

Symptoms: Error: Element is not attached to the DOM Detection Rule: If you see detachment errors during dynamic content updates Fix: Replace element variables with fresh locator calls:

// BAD: Storing element reference
const button = page.locator('button');
await button.click(); // May fail if DOM updated

// GOOD: Fresh locator each time
await page.locator('button').click();

Timeout Hell

Symptoms: Tests fail with "Timeout exceeded" in CI but pass locally Detection Rule: If tests have inconsistent CI failures with 30s+ timeouts Fix: Implement explicit waits with proper conditions:

// BAD: Blind timeout increase
await page.waitForTimeout(5000);

// GOOD: Wait for specific condition
await page.waitForSelector('[data-testid="loading"]', { state: 'hidden' });
await expect(page.locator('[data-testid="results"]')).toBeVisible();

Race Condition Rapids

Symptoms: Intermittent failures where elements "aren't ready yet" Detection Rule: If test flakiness correlates with slow network/CPU Fix: Chain waits to ensure proper sequencing:

// BAD: Assuming immediate availability
await page.click('#submit');
await page.fill('#new-field', 'value'); // May fail

// GOOD: Wait for UI state transition
await page.click('#submit');
await page.waitForSelector('#new-field:not([disabled])');
await page.fill('#new-field', 'value');

Selector Fragility Syndrome

Symptoms: Tests break when CSS classes or DOM structure changes Detection Rule: If tests fail after frontend refactoring without feature changes Fix: Migrate to semantic locators:

// BAD: Structural dependency
await page.click('.header > .nav > .item:nth-child(3)');

// GOOD: Semantic meaning
await page.getByRole('navigation').getByRole('link', { name: 'Products' }).click();

Screenshot Drift Disease

Symptoms: Visual regression tests fail due to minor rendering differences Detection Rule: If screenshot tests fail in CI with <5% pixel differences Fix: Configure appropriate tolerances and masks:

await expect(page).toHaveScreenshot('page.png', {
  maxDiffPixelRatio: 0.01,
  mask: [page.locator('[data-testid="dynamic-timestamp"]')],
  animations: 'disabled'
});

Worked Examples

Authentication Flow Test Implementation

Scenario: Test user login with error handling and success validation

Expert Approach:

import { test, expect } from '@playwright/test';

test.describe('User Authentication', () => {
  test('should handle complete login flow', async ({ page }) => {
    // Decision: Use Page Object for multi-step flow
    const loginPage = new LoginPage(page);
    
    await loginPage.goto();
    
    // Decision: Test error state first (negative case)
    await loginPage.signIn('invalid@email.com', 'wrongpass');
    
    // Wait strategy: Error message should appear
    await expect(page.getByRole('alert')).toContainText('Invalid credentials');
    
    // Decision: Clear state before positive test
    await loginPage.clearForm();
    
    // Decision: Use valid test data
    await loginPage.signIn('user@example.com', 'validpass');
    
    // Wait strategy: Navigation indicates success
    await page.waitForURL('/dashboard');
    
    // Verification: Check authenticated state
    await expect(page.getByRole('heading', { name: 'Dashboard' })).toBeVisible();
    await expect(page.getByText('Welcome back, John')).toBeVisible();
  });
});

class LoginPage {
  constructor(private page: Page) {}
  
  async goto() {
    await this.page.goto('/login');
    // Wait for form to be interactive
    await this.page.waitForSelector('form[data-testid="login-form"]');
  }
  
  async signIn(email: string, password: string) {
    // Decision: Use semantic locators
    await this.page.getByLabel('Email').fill(email);
    await this.page.getByLabel('Password').fill(password);
    await this.page.getByRole('button', { name: 'Sign In' }).click();
  }
  
  async clearForm() {
    await this.page.getByLabel('Email').clear();
    await this.page.getByLabel('Password').clear();
  }
}

Novice would miss: Error state testing, proper wait strategies, form state management Expert catches: Complete flow coverage, semantic locators, defensive waits

Quality Gates

Test implementation checklist:

• All user interactions use semantic locators (getByRole, getByLabel, getByText)
• No hard waits (page.waitForTimeout) - only conditional waits
• Error states and edge cases are tested, not just happy path
• Tests pass consistently in CI (5+ consecutive runs without flakes)
• Each test is independent and can run in any order
• Page Object Model used for multi-step flows (3+ interactions)
• Visual regression tests have appropriate tolerance thresholds (<2% pixel diff)
• API mocking implemented for external dependencies
• Authentication state properly managed via fixtures
• Test coverage includes critical user journeys (login, checkout, etc.)

NOT-FOR Boundaries

Don't use this skill for:

• Unit testing individual functions → Use vitest-testing-patterns instead
• API contract testing → Use api-architect for endpoint validation
• Performance testing → Use dedicated load testing tools like k6
• Security penetration testing → Use specialized security testing tools
• Cross-browser compatibility issues → Use browser-specific debugging tools

Delegate to other skills:

• For CI/CD pipeline setup → Use github-actions-pipeline-builder
• For accessibility auditing → Use accessibility-auditor
• For component testing → Use vitest-testing-patterns