A suite of tools for running automated end-to-end (E2E) tests using Playwright and performing AI-driven self-healing on detected errors.
This skill provides a robust framework for testing the tsgabrielle® web application. It automates browser interactions (visiting pages, clicking buttons, filling forms) to verify functionality and uses AI analysis to diagnose and potentially fix issues.
Ensure Playwright is installed in the workspace.
npm install -D @playwright/test
npx playwright install chromium --with-deps
When the user asks to "test" a specific feature (e.g., "Test the checkout flow" or "Verify the product page loads"), do not manually browse. Instead:
e2e/[feature-name].spec.ts using the template below.getByRole, getByText, or getByLabel. Avoid rigid CSS selectors if possible.import { test, expect } from '@playwright/test';
test('User Flow: [Feature Name]', async ({ page }) => {
// 1. Visit URL
await page.goto('http://localhost:3000/[route]');
// 2. Interaction Steps
await page.getByRole('button', { name: 'Add to Cart' }).click();
// 3. Assertions
await expect(page.getByText('Added to Bag')).toBeVisible();
});
Execute the tests using the helper script or direct CLI.
Command:
npx playwright test e2e/[feature-name].spec.ts
Options:
--headed: Run with visible browser (useful for debugging).--debug: Run in inspector mode.If a test fails:
grep_search or view_file to find the component code related to the failing test.await expect(...).toBeVisible() wait steps.baseURL is set to localhost).Generated by Antigravity AutoTester Skill