E2e Testing

Constraints

• Apply @rules/code-testing/general.mdc for the deterministic-test contract: tests must be reliable, isolated, and free of arbitrary sleeps.
• GATED skill: proceed only when the consuming project already ships Playwright. Never install Playwright unprompted.
• Stable selectors over brittle ones: prefer role and data-testid over CSS/structure.
• No arbitrary waitForTimeout. Rely on auto-waiting locators and explicit conditions.

Use when

• Writing Playwright E2E browser tests for a Laravel app.
• Stabilizing flaky Playwright tests.
• Wiring Playwright into CI against the app under test.

Preconditions (run first, do not skip)

Check whether Playwright is actually present:

ls playwright.config.* 2>/dev/null
grep -q '@playwright/test' package.json 2>/dev/null && echo "playwright present"

• If neither a playwright.config.* file nor @playwright/test in package.json exists, STOP. Do not install Playwright. Defer to:
• @skills/test-like-human/SKILL.md for manual, scenario-based testing, or
• Pest feature tests / Laravel Dusk for browser coverage inside the PHP stack.
• Only when Playwright IS present, continue with the patterns below.

File organization

tests/
  e2e/
    auth/login.spec.ts
    features/search.spec.ts
  pages/
    ItemsPage.ts
  fixtures/
    auth.ts
playwright.config.ts

Page Object Model

Encapsulate page structure so specs read as behavior, not selectors.

import { Page, Locator } from '@playwright/test'

export class ItemsPage {
  readonly page: Page
  readonly searchInput: Locator
  readonly itemCards: Locator
  readonly createButton: Locator

  constructor(page: Page) {
    this.page = page
    // Prefer role/test-id selectors. data-testid is stable across restyles.
    this.searchInput = page.getByTestId('search-input')
    this.itemCards = page.getByTestId('item-card')
    this.createButton = page.getByRole('button', { name: 'Create' })
  }

  async goto() {
    await this.page.goto('/items')
  }

  async search(query: string) {
    await this.searchInput.fill(query)
    // Wait for the real response, not a fixed delay.
    await this.page.waitForResponse(r => r.url().includes('/items') && r.ok())
  }

  itemCount() {
    return this.itemCards.count()
  }
}

Stable selectors

Order of preference:

1. getByRole(...) with an accessible name.
2. getByTestId(...) — add data-testid to the relevant Blade/Livewire markup.
3. getByLabel / getByText for forms and copy.
4. Raw CSS only as a last resort; it breaks on Tailwind restyles.

Test structure

import { test, expect } from '@playwright/test'
import { ItemsPage } from '../pages/ItemsPage'

test.describe('Item search', () => {
  let items: ItemsPage

  test.beforeEach(async ({ page }) => {
    items = new ItemsPage(page)
    await items.goto()
  })

  test('searches by keyword', async () => {
    await items.search('test')
    expect(await items.itemCount()).toBeGreaterThan(0)
    await expect(items.itemCards.first()).toContainText(/test/i)
  })

  test('shows empty state with no results', async ({ page }) => {
    await items.search('zzz-nonexistent')
    await expect(page.getByTestId('no-results')).toBeVisible()
    expect(await items.itemCount()).toBe(0)
  })
})

Running against the Laravel app

Point Playwright at the running app and let it boot php artisan serve (or use an already-running instance / CI app URL).

import { defineConfig, devices } from '@playwright/test'

export default defineConfig({
  testDir: './tests/e2e',
  fullyParallel: true,
  forbidOnly: !!process.env.CI,
  retries: process.env.CI ? 2 : 0,
  reporter: [['html', { outputFolder: 'playwright-report' }], ['junit', { outputFile: 'playwright-results.xml' }]],
  use: {
    baseURL: process.env.APP_URL ?? 'http://127.0.0.1:8000',
    trace: 'on-first-retry',
    screenshot: 'only-on-failure',
    video: 'retain-on-failure',
  },
  projects: [
    { name: 'chromium', use: { ...devices['Desktop Chrome'] } },
    { name: 'mobile-chrome', use: { ...devices['Pixel 5'] } },
  ],
  webServer: {
    command: 'php artisan serve --port=8000',
    url: 'http://127.0.0.1:8000',
    reuseExistingServer: !process.env.CI,
    timeout: 120_000,
  },
})

Use a dedicated testing database/env (.env.testing) so E2E runs never touch development data, and migrate/seed it before the suite.

Auth and network state reuse

Log in once, save storage state, and reuse it so specs skip the login flow.

// global-setup.ts — sign in once, persist cookies/session
import { chromium } from '@playwright/test'

export default async () => {
  const browser = await chromium.launch()
  const page = await browser.newPage()
  await page.goto(`${process.env.APP_URL ?? 'http://127.0.0.1:8000'}/login`)
  await page.getByLabel('Email').fill(process.env.E2E_EMAIL!)
  await page.getByLabel('Password').fill(process.env.E2E_PASSWORD!)
  await page.getByRole('button', { name: 'Log in' }).click()
  await page.waitForURL('**/dashboard')
  await page.context().storageState({ path: 'tests/e2e/.auth/user.json' })
  await browser.close()
}

Wire it via globalSetup and use: { storageState: 'tests/e2e/.auth/user.json' }. Keep .auth/ out of version control. You can also stub backend responses with page.route(...) to isolate the frontend from flaky external services.

CI wiring

name: E2E
on: [push, pull_request]
jobs:
  e2e:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: shivammathur/setup-php@v2
        with: { php-version: '8.3' }
      - run: composer install --no-interaction --prefer-dist
      - uses: actions/setup-node@v4
        with: { node-version: 20 }
      - run: npm ci
      - run: npx playwright install --with-deps
      - run: npm run build
      - run: php artisan migrate --seed --env=testing
      - run: npx playwright test
        env:
          APP_URL: http://127.0.0.1:8000
      - uses: actions/upload-artifact@v4
        if: always()
        with: { name: playwright-report, path: playwright-report/, retention-days: 30 }

Flaky-test avoidance

Per @rules/code-testing/general.mdc, tests must be deterministic.

• Use auto-waiting locators (locator.click()), never page.click(selector) on a maybe-not-ready node.
• Replace await page.waitForTimeout(...) with waitForResponse, waitForURL, or expect(locator).toBeVisible().
• Wait for the element state before acting on animated UI: await locator.waitFor({ state: 'visible' }).
• Quarantine genuinely flaky tests with test.fixme(true, 'flaky — issue #NNN') and link an issue rather than leaving them red.
• Reproduce flakiness with npx playwright test path.spec.ts --repeat-each=10.

Done when

• Preconditions confirmed Playwright is present (otherwise deferred to manual/Pest/Dusk).
• Specs use POM and role/test-id selectors.
• The suite runs against the Laravel app on a dedicated testing env/database.
• Auth/network state is reused where it speeds the suite.
• No arbitrary sleeps; all waits are condition-based.
• CI runs the suite and uploads the report.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: pekral
• Source: pekral/cursor-rules
• License: MIT
• Homepage: https://pekral.cz

Install and usage instructions live in the source repository linked above.