⚡ What’s Inside
- → Setup commands, locators (
getByRole,getByText,getByLabel), assertions, and page actions - → Claude AI prompts for generating Playwright tests automatically
- → All code is TypeScript/JavaScript, copy-paste ready — use Ctrl+F / Cmd+F to jump to the section you need
1 Playwright Setup Commands
Every command you need to initialize, run, and debug Playwright tests. For every flag and filter, see how to run Playwright tests.
| Command | What it does |
|---|---|
| npm init playwright@latest | Scaffold a new Playwright project with config, example tests, and browsers |
| npx playwright test | Run all tests across all configured projects |
| npx playwright test --ui | Open the interactive UI Mode for watching and debugging tests |
| npx playwright codegen url | Launch Codegen — record browser actions and generate test code |
| npx playwright show-report | Open the HTML test report in your default browser |
| npx playwright test --debug | Run tests with the Playwright Inspector (step-through debugger) |
| npx playwright test --project=chromium | Run tests on Chromium only (skip Firefox and WebKit) |
| npx playwright test file.spec.ts | Run a single test file |
Tip: Add --headed to any test command to watch the browser run visually. Combine with --workers=1 to slow things down for debugging.
2 Essential Locators
Playwright recommends user-facing locators first. Fall back to CSS/XPath only when necessary.
| Locator | Use case |
|---|---|
| page.getByRole('button', { name: 'Submit' }) | Best practice — matches accessible role + name |
| page.getByText('Welcome') | Find element by visible text content |
| page.getByLabel('Email') | Target form inputs by their associated label |
| page.getByPlaceholder('Enter email') | Target inputs by placeholder text |
| page.getByTestId('login-form') | Use data-testid attribute (stable, refactor-safe) |
| page.locator('.class-name') | CSS selector — use when semantic locators aren't available |
| page.locator('#id') | ID selector — fast but brittle if IDs are auto-generated |
| page.locator('css=div >> text=hello') | Chained locator — CSS parent with text child filter |
3 Common Assertions
All assertions auto-wait and retry until timeout. No sleep() needed.
// Page-level assertions await expect(page).toHaveTitle(/Playwright/); await expect(page).toHaveURL(/dashboard/); // Element visibility & text await expect(locator).toBeVisible(); await expect(locator).toHaveText('Hello'); await expect(locator).toBeEnabled(); // Count & attributes await expect(locator).toHaveCount(3); await expect(locator).toHaveAttribute('href', '/about'); // Visual regression await expect(locator).toHaveScreenshot();
Tip: Negate any assertion with .not — e.g. await expect(locator).not.toBeVisible();
4 Page Actions
The most-used page interactions, copy-paste ready.
// Navigation await page.goto('https://example.com'); // Click & fill await page.click('button'); await page.fill('#email', 'test@test.com'); // Form interactions await page.selectOption('select', 'value'); await page.check('#agree'); // Waiting await page.waitForSelector('.loaded'); // Screenshots await page.screenshot({ path: 'screenshot.png' }); // Execute JS in browser context await page.evaluate(() => document.title);
5 Claude AI + MCP Prompts
Real prompts you can paste into Claude to generate production-ready Playwright tests instantly.
"Write a Playwright test that logs in with email/password and verifies the dashboard loads"
"Generate a POM class for the checkout page with methods for addToCart, applyCoupon, and placeOrder"
"Create a data-driven test that validates form validation for 5 invalid email formats"
"Write an API test that creates a user via POST, then verifies the user appears in the UI"
"Fix this broken locator: [paste locator] — the button text changed from 'Submit' to 'Send'"
How it works: The Playwright MCP Server connects Claude directly to your browser. Claude can navigate pages, inspect elements, and generate tests based on the live DOM — not just guessing from documentation. Learn the full MCP setup here.
6 playwright.config.ts Quick Reference
A production-ready config with the most common options annotated.
import { defineConfig, devices } from '@playwright/test'; export default defineConfig({ testDir: './tests', timeout: 30_000, // per-test timeout (ms) retries: 2, // retry failed tests on CI workers: '50%', // use half of CPU cores fullyParallel: true, // parallelize within files reporter: [ ['html', { open: 'never' }], // HTML report ['allure-playwright'], // Allure (optional) ], use: { baseURL: 'http://localhost:3000', trace: 'on-first-retry', // capture trace on failures screenshot: 'only-on-failure', video: 'retain-on-failure', // save video only if test fails }, projects: [ { name: 'chromium', use: { ...devices['Desktop Chrome'] }, }, { name: 'firefox', use: { ...devices['Desktop Firefox'] }, }, { name: 'webkit', use: { ...devices['Desktop Safari'] }, }, { name: 'mobile-chrome', use: { ...devices['Pixel 7'] }, }, { name: 'mobile-safari', use: { ...devices['iPhone 14'] }, }, ], });
Asim Noaman
Sr. SQA Manager & AI Automation Engineer
18+ years in QA. Asim currently runs an 80+ engineer QA function at Tkxel and built a Testing Center of Excellence for a major Saudi bank. He builds AI testing agents with Playwright, Claude AI and the MCP Server at work, and teaches the same workflow in his courses.
Playwright + Claude AI Course
Take Your Playwright Skills Further With AI
This guide covers one piece of the puzzle. The course teaches Playwright in JavaScript from scratch (locators, assertions, hooks and the Page Object Model), then Claude Code, the Playwright MCP Server, your own QA AI agent and Playwright Test Agents.
- Everything in this guide, plus the complete Playwright stack
- Claude AI generates tests 3–5× faster than writing them by hand
- Claude Code and the Playwright MCP Server for AI test generation
- Build your own QA AI agent with Claude and MCP