Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Workflows für Playwright, Puppeteer, Web Scraping, E2E-Testing, Screenshots.
.claude/skills/shadd0wtaka-browser-automation-skill/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -31% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 48% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 46% | 0% |
| case-18 | ✗→✓ | ▲ Improved | 29% | 0% |
| case-19 | ✗→✓ | ▲ Improved | 2% | 0% |
Workflows für Playwright, Puppeteer, Web Scraping, E2E-Testing, Screenshots.
bashnpm init playwright@latest npx playwright install chromium # + firefox, webkit # Oder Docker: docker run --rm --network host mcr.microsoft.com/playwright:v1.45.0 npx playwright test
pythonfrom playwright.sync_api import sync_playwright with sync_playwright() as p: browser = p.chromium.launch(headless=True) page = browser.new_page(viewport={"width": 1280, "height": 720}) page.goto("https://example.com") page.fill("input[name=q]", "search term") page.click("button[type=submit]") page.wait_for_selector(".results") html = page.content() # Screenshot page.screenshot(path="screenshot.png", full_page=True) # PDF page.pdf(path="page.pdf") browser.close()
typescriptimport { test, expect } from '@playwright/test'; test('user can login', async ({ page }) => { await page.goto('/login'); await page.fill('#email', 'user@example.com'); await page.fill('#password', 'pass123'); await page.click('button[type="submit"]'); await expect(page.locator('.welcome')).toBeVisible(); await expect(page).toHaveURL(/\/dashboard/); });
javascriptimport puppeteer from 'puppeteer'; const browser = await puppeteer.launch({ headless: 'new' }); const page = await browser.newPage(); await page.goto('https://example.com', { waitUntil: 'networkidle0' }); const text = await page.evaluate(() => document.body.innerText); await page.screenshot({ path: 'page.png' }); await browser.close();
python# Combine browser + AI for intelligent scraping page.goto("https://docs.example.com") content = page.inner_text("main") # Analyze via OmniRoute resp = httpx.post( "http://localhost:20128/v1/chat/completions", json={ "model": "oc/deepseek-v4-flash-free", "messages": [ {"role": "system", "content": "Extract all API endpoints from this documentation."}, {"role": "user", "content": content[:8000]}, ], }, )
bash# Visual Regression npx playwright test --update-snapshots # update baselines npx playwright test # compare # Component Testing (Storybook + Playwright) npx test-storybook --url http://localhost:6006 # Accessibility npx playwright test --project=accessibility await page.accessibility.snapshot()
bash# Slow motion + DevTools page = browser.new_page(slow_mo=500, devtools=True) # Trace await page.tracing.start({screenshots: true, snapshots: true}) # ... test ... await page.tracing.stop(path="trace.zip") # View: https://trace.playwright.dev/
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-02 | fail→fail | 17,411 | 14,805 | -15% | 1 | 1 | 0% | 3,303 | 3,561 | +8% | 0 | 0 | — |
case-01 | fail→fail | 6,654 | 5,488 | -18% | 1 | 1 | 0% | 1,330 | 1,917 | +44% | 0 | 0 | — |
case-03 | pass→pass | 7,742 | 6,518 | -16% | 1 | 1 | 0% | 1,597 | 2,168 | +36% | 0 | 0 | — |
case-04 | fail→pass | 12,118 | 3,850 | -68% | 1 | 1 | 0% | 2,235 | 1,536 | -31% | 0 | 0 | — |
case-05 | fail→pass | 7,233 | 5,979 | -17% | 1 | 1 | 0% | 1,387 | 2,054 | +48% | 0 | 0 | — |
case-06 | pass→pass | 4,233 | 1,696 | -60% | 1 | 1 | 0% | 725 | 1,029 | +42% | 0 | 0 | — |
case-07 | pass→pass | 13,162 | 7,159 | -46% | 1 | 1 | 0% | 2,391 | 2,024 | -15% | 0 | 0 | — |
case-08 | pass→pass | 5,751 | 1,502 | -74% | 1 | 1 | 0% | 944 | 990 | +5% | 0 | 0 | — |
case-09 | fail→fail | 11,025 | 9,219 | -16% | 1 | 1 | 0% | 2,152 | 2,625 | +22% | 0 | 0 | — |
case-10 | fail→fail | 7,863 | 5,980 | -24% | 1 | 1 | 0% | 1,488 | 2,017 | +36% | 0 | 0 | — |
case-11 | pass→pass | 3,779 | 2,849 | -25% | 1 | 1 | 0% | 651 | 1,239 | +90% | 0 | 0 | — |
case-12 | pass→pass | 5,762 | 3,375 | -41% | 1 | 1 | 0% | 901 | 1,323 | +47% | 0 | 0 | — |
case-13 | pass→pass | 3,834 | 2,744 | -28% | 1 | 1 | 0% | 624 | 1,260 | +102% | 0 | 0 | — |
case-14 | pass→pass | 2,399 | 2,129 | -11% | 1 | 1 | 0% | 392 | 1,151 | +194% | 0 | 0 | — |
case-15 | pass→pass | 7,433 | 4,344 | -42% | 1 | 1 | 0% | 1,293 | 1,517 | +17% | 0 | 0 | — |
case-16 | fail→pass | 5,909 | 4,725 | -20% | 1 | 1 | 0% | 1,131 | 1,653 | +46% | 0 | 0 | — |
case-17 | pass→pass | 6,573 | 4,162 | -37% | 1 | 1 | 0% | 1,137 | 1,501 | +32% | 0 | 0 | — |
case-18 | fail→pass | 6,019 | 3,731 | -38% | 1 | 1 | 0% | 1,049 | 1,353 | +29% | 0 | 0 | — |
case-19 | fail→pass | 15,841 | 11,533 | -27% | 1 | 1 | 0% | 2,749 | 2,809 | +2% | 0 | 0 | — |
case-20 | pass→pass | 4,340 | 4,550 | +5% | 1 | 1 | 0% | 793 | 1,635 | +106% | 0 | 0 | — |
case-21 | pass→pass | 4,715 | 3,676 | -22% | 1 | 1 | 0% | 820 | 1,388 | +69% | 0 | 0 | — |
case-22 | pass→pass | 9,141 | 8,679 | -5% | 1 | 1 | 0% | 1,842 | 2,536 | +38% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +23 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.