Skip to content
Back to skills

Webapp Testing

ASecurity

Browser-based UI verification using Playwright. Page Object Model, selector best practices, visual regression, network interception, and MCP integration. Trigger: When writing E2E tests, verifying UI changes, or setting up Playwright.

  • 5 stars
  • 0 votes
  • 0 copies
  • 1 view
  • Added September 26, 2026
ai-agentstypescriptgobashtestingapi

Works with

  • cli
  • api
  • mcp

Security analysis

A100/100

Scanned September 26, 2026

npx -y skills add JNZader/repoforge --skill webapp-testing --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Webapp Testing?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Webapp Testing
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/jnzader-webapp-testing/badge)](https://www.skillsdirectory.com/skills/jnzader-webapp-testing)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: webapp-testing
description: >
  Browser-based UI verification using Playwright. Page Object Model, selector
  best practices, visual regression, network interception, and MCP integration.
  Trigger: When writing E2E tests, verifying UI changes, or setting up Playwright.
license: MIT
metadata:
  author: JNZader
  version: "1.0"
  tags: [playwright, e2e, testing, ui, browser]
  category: testing
  source: anthropics/skills
allowed-tools: Read, Write, Edit, Bash, Glob, Grep
---

## Purpose

Patterns for browser-based UI verification using Playwright. Reliable, maintainable, fast E2E tests.

---

## Selector Priority

| Priority | Selector | Example | Why |
|----------|----------|---------|-----|
| 1 | Role | `getByRole('button', { name: 'Submit' })` | Accessible, semantic |
| 2 | Label | `getByLabel('Email')` | Accessible, user-facing |
| 3 | Test ID | `getByTestId('submit-btn')` | Stable, explicit |
| 4 | Text | `getByText('Sign in')` | Readable but fragile |
| 5 | CSS | `page.locator('.btn-primary')` | Last resort |

NEVER: XPath, auto-generated IDs, DOM structure selectors.

---

## Page Object Model

```typescript
export class LoginPage {
  constructor(private page: Page) {}
  get emailInput() { return this.page.getByLabel('Email'); }
  get passwordInput() { return this.page.getByLabel('Password'); }
  get submitButton() { return this.page.getByRole('button', { name: 'Sign in' }); }

  async login(email: string, password: string) {
    await this.emailInput.fill(email);
    await this.passwordInput.fill(password);
    await this.submitButton.click();
  }
}
```

---

## Visual Regression

```typescript
test('dashboard renders correctly', async ({ page }) => {
  await page.goto('/dashboard');
  await page.waitForLoadState('networkidle');
  await expect(page).toHaveScreenshot('dashboard.png', { maxDiffPixelRatio: 0.01 });
});
```

---

## Network Interception

```typescript
test('handles API errors', async ({ page }) => {
  await page.route('**/api/users', route =>
    route.fulfill({ status: 500, body: 'Server Error' })
  );
  await page.goto('/users');
  await expect(page.getByText('Something went wrong')).toBeVisible();
});
```

---

## Anti-Flakiness

1. Wait for state, not time — never `page.waitForTimeout()`
2. Isolate tests — fresh context per test
3. Web-first assertions — `expect(locator).toBeVisible()` auto-retries
4. Mock external APIs
5. CI retries: `retries: 2` in config

---

## Critical Rules

1. NEVER use `page.waitForTimeout()` — wait for specific conditions
2. ALWAYS use Page Object Model for pages with 3+ interactions
3. Selectors MUST follow priority: role > label > testid > text > CSS
4. Visual regression MUST set `maxDiffPixelRatio`
5. E2E tests MUST NOT depend on seed data
6. ALWAYS use `--trace on-first-retry` in CI

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.