Web Application Testing

To test local web applications, write native Python Playwright scripts.

Helper Scripts Available:

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

Always run scripts with --help first to see usage. DO NOT read the source until you try running the script first and find that a customized solution is abslutely necessary. These scripts can be very large and thus pollute your context window. They exist to be called directly as black-box scripts rather than ingested into your context window.

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 → 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

Example: Using with_server.py

To start a server, run --help first, then use the helper:

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

To create an automation script, include only Playwright logic (servers are managed automatically):

from playwright.sync_api import sync_playwright

with sync_playwright() as p:
    browser = p.chromium.launch(headless=True) # Always launch chromium in headless mode
    page = browser.new_page()
    page.goto('http://localhost:5173') # Server already running and ready
    page.wait_for_load_state('networkidle') # CRITICAL: Wait for JS to execute
    # ... your automation logic
    browser.close()

Reconnaissance-Then-Action Pattern

1. Inspect rendered DOM:

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

2. Identify selectors from inspection results

3. Execute actions using discovered selectors

Common Pitfall

❌ Don't inspect the DOM before waiting for networkidle on dynamic apps ✅ Do wait for page.wait_for_load_state('networkidle') before inspection

Best Practices

• Use bundled scripts as black boxes - To accomplish a task, consider whether one of the scripts available in scripts/ can help. These scripts handle common, complex workflows reliably without cluttering the context window. Use --help to see usage, then invoke 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()

Reference Files

• examples/ - Examples showing common patterns:
  • element_discovery.py - Discovering buttons, links, and inputs on a page
  • static_html_automation.py - Using file:// URLs for local HTML
  • console_logging.py - Capturing console logs during automation
  • node_site_diagnostics.js - Simple Node diagnostics (console errors/HTTP failures collection + screenshots)

Node Playwright Addendum (local extensions)

Documenting useful patterns from Node operations. The official content above remains Python-based; this section serves as a local extension for reference.

• Quick one-liner: Fastest approach without cluttering /tmp. Minimal example for networkidle wait and full-page screenshot:

  node -e "const { chromium } = require('playwright');
  (async () => {
    const browser = await chromium.launch();
    const page = await browser.newPage();
    await page.goto(process.env.BASE_URL || 'http://localhost:3000', { waitUntil: 'networkidle' });
    await page.screenshot({ path: '/tmp/webapp.png', fullPage: true });
    await browser.close();
    console.log('saved: /tmp/webapp.png');
  })();"
  

• Evidence collection (scripts/videos/screenshots/traces): When evidence is required, consolidate in .artifacts/<feature>/. Save the Playwright script itself in scripts/ to make execution reproducible. Videos use recordVideo, traces use --trace=retain-on-failure in Playwright Test for convenience.

  FEATURE=${FEATURE:-feature}
  mkdir -p .artifacts/$FEATURE/{scripts,images,videos}
  node -e "const { chromium } = require('playwright');
  (async () => {
    const browser = await chromium.launch({ headless: false });
    const context = await browser.newContext({
      viewport: { width: 1440, height: 900 },
      recordVideo: { dir: `.artifacts/${FEATURE}/videos` }
    });
    const page = await context.newPage();
    await page.goto(process.env.BASE_URL || 'http://localhost:3000', { waitUntil: 'networkidle' });
    await page.screenshot({ path: `.artifacts/${FEATURE}/images/${Date.now()}-step.png`, fullPage: true });
    await browser.close();
  })();"
  # When keeping traces with Playwright Test
  # BASE_URL=http://localhost:3000 npx playwright test tests/e2e/<spec>.spec.ts --headed --output=.artifacts/$FEATURE/images --trace=retain-on-failure --reporter=line
  

• Decision on Chrome DevTools MCP usage: For layout/font/stacking/performance issues hard to diagnose from screenshots alone, first reproduce and capture evidence with Playwright → if still unclear, pinpoint with DevTools MCP for Styles/Computed/Box model/Performance.

• Performance snapshot with Lighthouse: Minimal execution for rough performance measurement. Output consolidated in /tmp.

  npx lighthouse ${BASE_URL:-http://localhost:3000} --output=json --output-path=/tmp/lh.json --chrome-flags="--headless" --only-categories=performance
  node -e "const data = require('/tmp/lh.json'); const perf = data.categories.performance; console.log('Performance Score', Math.round(perf.score*100));"
  

Operation policy: Default to headless, switch to headed only when evidence is needed. Write to /tmp or .artifacts/ subdirectory without cluttering the project root, and delete unnecessary files after completion.

Notes on Playwright-only approach without DevTools MCP

Chrome DevTools MCP internally uses Puppeteer + CDP. Playwright can also use CDP, so substitute with the following steps.

• Performance trace (equivalent to Performance panel): Use Playwright's standard tracing.

  with sync_playwright() as p:
      browser = p.chromium.launch(headless=True)
      context = browser.new_context(record_video_dir=None)
      context.tracing.start(screenshots=True, snapshots=True)
      page = context.new_page()
      page.goto("http://localhost:3000", wait_until="networkidle")
      # Perform operations here
      context.tracing.stop(path=".artifacts/feature/traces/trace.zip")
      browser.close()
  

  For detailed output similar to DevTools Performance view, use a CDP session with Tracing.start/end and read the output JSON with chrome://tracing or perfetto.dev.

• Coverage (equivalent to Coverage panel): Retrieve via CDP.

  cdp = page.context.new_cdp_session(page)
  cdp.send("Profiler.enable")
  cdp.send("Profiler.startPreciseCoverage", {"callCount": True, "detailed": True})
  # Perform operations here
  result = cdp.send("Profiler.takePreciseCoverage")
  cdp.send("Profiler.stopPreciseCoverage"); cdp.send("Profiler.disable")
  # result["result"] contains usage per file
  

• Checking Styles/Box Model/Computed values: Retrieve values without DevTools UI.

  box = page.locator("selector").evaluate("el => el.getBoundingClientRect()")
  styles = page.locator("selector").evaluate("el => getComputedStyle(el)")
  

• Network body retrieval: page.on('request') captures metadata, but response body needs CDP.

  cdp = page.context.new_cdp_session(page)
  resp = cdp.send("Network.getResponseBody", {"requestId": "<target requestId>"})
  

  Get requestId by logging with page.on("requestfinished", ...) along with request.timing() for correlation.

• Console/error collection: Playwright events are sufficient.

  page.on("console", lambda msg: print("console:", msg.type, msg.text))
  page.on("pageerror", lambda err: print("pageerror:", err))
  page.on("requestfailed", lambda req: print("requestfailed:", req.url))
  

File placement conventions

Consolidate verification-generated files in .artifacts/<feature>/ with the following structure:

.artifacts/<feature>/
├── scripts/      # Playwright scripts (.py / .js / .ts)
├── images/       # Screenshots
├── videos/       # Recorded videos
└── traces/       # Playwright traces (.zip)

• Scripts are also part of evidence: Save in scripts/ even if disposable, to make execution reproducible
• Naming convention: Use descriptive names like <timestamp>-<step>.png, verify-<feature>.py that convey intent
• Quick validation only in /tmp: OK for one-off checks without evidence, but assume no future reference