goose/documentation/docs/mcp/playwright-mcp.md

7.1 KiB

title description
Playwright Extension Add Playwright MCP Server as a goose Extension for Modern Web Testing

import Tabs from '@theme/Tabs'; import TabItem from '@theme/TabItem'; import YouTubeShortEmbed from '@site/src/components/YouTubeShortEmbed'; import GooseDesktopInstaller from '@site/src/components/GooseDesktopInstaller'; import CLIExtensionInstructions from '@site/src/components/CLIExtensionInstructions';

This tutorial covers how to add the Playwright MCP Server as a goose extension, to enable cross-browser testing and web automation across Chromium and Webkit.

:::tip Quick Install Launch the installer Command

npx -y @playwright/mcp@latest
:::

Configuration

:::info Note that you'll need Node.js installed on your system to run this command, as it uses npx. :::

Example Usage

Let's use goose with the Playwright extension to create a cross-browser testing suite. In this example, we'll demonstrate how goose can help you:

  1. Test a web application across multiple browsers
  2. Generate maintainable test code
  3. Capture screenshots for visual comparison

goose Prompt

Test the random redesign generator app (https://blackgirlbytes.github.io/random-redesign-picker/) 
in Chromium and WebKit. For each browser:
1. Navigate to the page
2. Click the "Pick My Combo" button
3. Verify the text changes
4. Take a screenshot
Then generate a Playwright test that works across all browsers.

goose Output

Let's start with Chromium:

─── browser_navigate | playwright ──────────────────────────
url: https://blackgirlbytes.github....

Now let's wait for and click the button:

─── browser_click | playwright ──────────────────────────
element: button with text "Pick My Combo"
ref: <button>

Let's capture the first combo and click again to verify it changes:

─── browser_click | playwright ──────────────────────────
element: button with text "Pick My Combo"
ref: <button>

Let's take a screenshot of the Chromium view:

─── browser_take_screenshot | playwright ──────────────────────────
filename: chromium-redesign.png

Now let's test in Safari. I'll open a new tab for Safari:

─── browser_tab_new | playwright ──────────────────────────
url: https://blackgirlbytes.github....

Let's click the button in Safari:

─── browser_click | playwright ──────────────────────────
element: button with text "Pick My Combo"
ref: <button>

Let's click again to verify the text changes:

─── browser_click | playwright ──────────────────────────
element: button with text "Pick My Combo"
ref: <button>

Let's take a screenshot of the Safari view:

─── browser_take_screenshot | playwright ──────────────────────────
filename: safari-redesign.png

Now I'll generate a Playwright test that runs in all three browser environments. The test will verify the core functionality we just tested manually:

─── browser_generate_playwright_test | playwright ──────────────────────────
name: Random Redesign Picker Test
description: Test the random redesign generator functionality across browsers
steps: [
  "Navigate to the application",
  "Verify initial page content",
  "Click button to generate first combination",
  "Click button again to generate second combination",
  "Verify combinations are different"
]

Here's the generated Playwright test that will run across Chromium, Firefox, and Safari:

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

test.describe('Random Redesign Picker', () => {
  test('should generate different random combinations', async ({ page }) => {
    // Navigate to the application
    await page.goto('https://blackgirlbytes.github.io/random-redesign-picker/');
    
    // Verify initial page content
    await expect(page).toHaveTitle('Redesign for the Ridiculous');
    await expect(page.getByRole('heading', { level: 1 })).toContainText('🎲 Redesign for the Ridiculous');
    
    // Verify button is visible
    const comboButton = page.getByRole('button', { name: '🎉 Pick My Combo' });
    await expect(comboButton).toBeVisible();
    
    // Click button and get first combination
    await comboButton.click();
    
    // Wait for and store first combo
    const firstComboElement = page.locator('generic').filter({ hasText: /^(?!.*🎲).*$/ }).last();
    await expect(firstComboElement).toBeVisible();
    const firstComboText = await firstComboElement.textContent();
    
    // Generate second combination
    await comboButton.click();
    
    // Wait for and get second combo
    const secondComboElement = page.locator('generic').filter({ hasText: /^(?!.*🎲).*$/ }).last();
    await expect(secondComboElement).toBeVisible();
    const secondComboText = await secondComboElement.textContent();
    
    // Verify combos are different
    expect(secondComboText).not.toEqual(firstComboText);
  });
});

To use this test, you would need to:

1. Save it as `tests/random-redesign.spec.ts`
2. Configure your `playwright.config.ts` to include all three browsers:

import { PlaywrightTestConfig } from '@playwright/test';

const config: PlaywrightTestConfig = {
  projects: [
    {
      name: 'chromium',
      use: { browserName: 'chromium' },
    },
    {
      name: 'firefox',
      use: { browserName: 'firefox' },
    },
    {
      name: 'webkit',
      use: { browserName: 'webkit' },
    },
  ],
};

export default config;

You can then run the tests using:

npx playwright test

This will execute the test in all three browser environments. The test verifies:
1. The page loads correctly
2. The title and heading are correct
3. The combo button is visible
4. Clicking the button generates a combination
5. Clicking again generates a different combination

The test uses robust selectors and proper waiting mechanisms to ensure reliability across different browsers and network conditions.