| name | webapp-testing |
| description | Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs. Activate on: Playwright, webapp testing, browser automation, E2E testing, UI testing. NOT for API-only testing without browser, unit tests, or mobile app testing. |
| allowed-tools | Read,Write,Edit,Bash,Glob,Grep |
| metadata | {"category":"Code Quality & Testing","pairs-with":[{"skill":"test-automation-expert","reason":"Comprehensive testing strategy"},{"skill":"site-reliability-engineer","reason":"Validate deployed web apps"}],"tags":["playwright","e2e","browser","automation","ui-testing"]} |
Web Application Testing
Write native Python Playwright scripts to test local web applications.
When to Use
✅ Use for:
- E2E testing of web applications
- UI automation and interaction testing
- Visual regression testing
- Browser log capture and debugging
- Screenshot capture for verification
- Form submission and validation testing
❌ NOT for:
- API-only testing without a browser (use requests/httpx)
- Unit testing of individual functions
- Mobile app testing (use Appium)
- Load/performance testing (use k6/Locust)
Decision Tree: Choosing Your Approach
User task → Is it static HTML?
├─ Yes → Read HTML file directly to identify selectors
│ ├─ Success → Write Playwright script using selectors
│ └─ Fails/Incomplete → Treat as dynamic (below)
│
└─ No (dynamic webapp) → Is the server already running?
├─ No → Start server first, then run Playwright
│
└─ Yes → Reconnaissance-then-action:
1. Navigate and wait for networkidle
2. Take screenshot or inspect DOM
3. Identify selectors from rendered state
4. Execute actions with discovered selectors
Core Playwright Patterns
Basic Test Structure
from playwright.sync_api import sync_playwright
with sync_playwright() as p:
browser = p.chromium.launch(headless=True)
page = browser.new_page()
page.goto('http://localhost:5173')
page.wait_for_load_state('networkidle')
browser.close()
Reconnaissance-Then-Action Pattern
Step 1: Inspect rendered DOM
page.screenshot(path='/tmp/inspect.png', full_page=True)
content = page.content()
buttons = page.locator('button').all()
Step 2: Identify selectors from inspection results
Step 3: Execute actions using discovered selectors
Selector Strategy (Priority Order)
-
Role-based (best for accessibility):
page.get_by_role("button", name="Submit")
page.get_by_role("textbox", name="Email")
-
Text-based (readable, but fragile to copy changes):
page.get_by_text("Sign In")
page.get_by_label("Password")
-
Test IDs (stable, explicit):
page.get_by_test_id("login-button")
-
CSS selectors (last resort):
page.locator(".btn-primary")
page.locator("#submit-form")
Common Anti-Patterns
Anti-Pattern: Not Waiting for Network Idle
Symptom: Tests pass locally, fail in CI; elements not found
Problem: Modern SPAs load content dynamically after initial page load
Solution:
page.goto('http://localhost:3000')
page.click('button')
page.goto('http://localhost:3000')
page.wait_for_load_state('networkidle')
page.click('button')
Anti-Pattern: Hardcoded Waits
Symptom: time.sleep(3) scattered throughout tests
Problem: Slow, unreliable, doesn't adapt to actual page state
Solution:
time.sleep(5)
page.click('.dynamic-button')
page.wait_for_selector('.dynamic-button', state='visible')
page.click('.dynamic-button')
Anti-Pattern: Inspecting DOM Before JavaScript Executes
Symptom: Empty page content, missing elements in static analysis
Problem: Reading HTML before client-side rendering completes
Solution: Always wait for networkidle on dynamic apps before inspection
Waiting Strategies
page.wait_for_selector('#my-element')
page.wait_for_selector('#my-element', state='visible')
page.wait_for_selector('#my-element', state='hidden')
page.wait_for_url('**/dashboard')
page.wait_for_load_state('networkidle')
page.wait_for_function('document.querySelector(".loaded")')
Screenshot Patterns
page.screenshot(path='/tmp/full.png', full_page=True)
page.locator('#header').screenshot(path='/tmp/header.png')
page.screenshot(path='/tmp/before.png')
page.screenshot(path='/tmp/after.png')
Console Log Capture
messages = []
page.on('console', lambda msg: messages.append({
'type': msg.type,
'text': msg.text
}))
page.on('console', lambda msg:
print(f'ERROR: {msg.text}') if msg.type == 'error' else None
)
Form Testing
page.fill('#email', 'test@example.com')
page.fill('#password', 'secret123')
page.select_option('#country', 'US')
page.check('#terms')
page.click('button[type="submit"]')
page.wait_for_url('**/success')
Assertions
from playwright.sync_api import expect
expect(page.locator('#title')).to_have_text('Welcome')
expect(page.locator('#count')).to_have_text('5')
expect(page.locator('.error')).to_be_hidden()
expect(page.locator('#submit')).to_be_enabled()
expect(page).to_have_url('http://localhost:3000/dashboard')
expect(page).to_have_title('My App')
Multi-Page Scenarios
with page.expect_popup() as popup_info:
page.click('#open-popup')
popup = popup_info.value
popup.wait_for_load_state()
with context.expect_page() as new_page_info:
page.click('a[target="_blank"]')
new_page = new_page_info.value
Test File Organization
tests/
├── conftest.py # Shared fixtures
├── test_login.py # Login flows
├── test_dashboard.py # Dashboard features
├── test_forms.py # Form submissions
└── screenshots/ # Visual artifacts
Running Tests
python -m pytest tests/test_login.py
PWDEBUG=1 python -m pytest tests/test_login.py
python -m pytest --tracing=on tests/test_login.py
Best Practices
- Use
sync_playwright() for synchronous scripts
- Always close the browser when done
- Use descriptive selectors: role, text, test-id over CSS
- Add appropriate waits:
wait_for_selector(), wait_for_load_state()
- Capture screenshots on failure for debugging
- Keep tests independent - each test should set up its own state
This skill encodes: Playwright best practices | Selector strategies | Wait patterns | Anti-pattern prevention | E2E testing workflows