Web Application Testing

Overview

This skill provides a toolkit for testing local web applications by writing native Python Playwright scripts. It includes a helper script that manages server lifecycles (including multi-server setups) and follows a reconnaissance-then-action pattern for reliably interacting with dynamic web content.

Helper Scripts Available:

• scripts/with_server.py -- Manages server lifecycle (supports multiple servers)

Always run scripts with --help first to see usage. Treat them as black-box utilities rather than reading their source.

Decision Tree

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 --> Run: python scripts/with_server.py --help
        |          Then use the helper + write simplified Playwright script
        |
        |-- 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

Using with_server.py

Single server:

python scripts/with_server.py --server "npm run dev" --port 5173 -- python your_automation.py

Multiple servers (e.g., backend + frontend):

python scripts/with_server.py \
  --server "cd backend && python server.py" --port 3000 \
  --server "cd frontend && npm run dev" --port 5173 \
  -- python your_automation.py

Writing the Automation Script

The server is managed by with_server.py, so the automation script only contains Playwright logic:

from playwright.sync_api import sync_playwright

with sync_playwright() as p:
    browser = p.chromium.launch(headless=True)  # Always headless
    page = browser.new_page()
    page.goto('http://localhost:5173')           # Server already running
    page.wait_for_load_state('networkidle')      # CRITICAL: wait for JS
    # ... automation logic here
    browser.close()

Reconnaissance-Then-Action Pattern

1. Inspect the rendered DOM:

   page.screenshot(path='/tmp/inspect.png', full_page=True)
   content = page.content()
   page.locator('button').all()
   

2. Identify selectors from the inspection results

3. Execute actions using the discovered selectors

Common Pitfall

• Wrong: Inspecting the DOM before waiting for networkidle on dynamic apps
• Right: Always call page.wait_for_load_state('networkidle') before inspection

Best Practices

• Use bundled scripts as black boxes -- invoke with --help then call directly
• Use sync_playwright() for synchronous scripts
• Always close the browser when done
• Use descriptive selectors: text=, role=, CSS selectors, or IDs
• Add appropriate waits: page.wait_for_selector() or page.wait_for_timeout()

Example Files