Puppeteer Browser Automation Skill

Automate Chrome/Chromium with Puppeteer — Google's Node.js library for Chrome DevTools Protocol.

Installation

npm install puppeteer          # Downloads Chromium (~170MB)
# or
npm install puppeteer-core     # No bundled browser (bring your own)

Core Patterns

Open a page

const puppeteer = require('puppeteer');

const browser = await puppeteer.launch({ headless: true });
const page = await browser.newPage();
await page.goto('https://example.com');
console.log(await page.title());
await browser.close();

Click, type, select

await page.click('button.submit');
await page.type('#username', 'my_user');        // Types character by character
await page.type('#password', 'my_pass');
await page.select('select#country', 'US');
await page.keyboard.press('Enter');

Wait for elements

await page.waitForSelector('.results', { timeout: 10000 });
await page.waitForNavigation({ waitUntil: 'networkidle0' });
await page.waitForFunction('document.querySelector(".loaded")');

Extract data

const title = await page.$eval('h1', el => el.textContent);
const href = await page.$eval('a.link', el => el.getAttribute('href'));
const items = await page.$$eval('.product', els =>
  els.map(el => ({
    name: el.querySelector('.name').textContent,
    price: el.querySelector('.price').textContent,
  }))
);

Screenshots and PDF

await page.screenshot({ path: 'page.png' });
await page.screenshot({ path: 'full.png', fullPage: true });
await page.pdf({ path: 'report.pdf', format: 'A4', printBackground: true });

Network interception

await page.setRequestInterception(true);
page.on('request', request => {
  if (request.resourceType() === 'image') {
    request.abort();
  } else {
    request.continue();
  }
});

Connect to existing browser

const browser = await puppeteer.connect({
  browserWSEndpoint: 'ws://127.0.0.1:9222/devtools/browser/xxx'
});

Set viewport and user agent

await page.setViewport({ width: 1920, height: 1080 });
await page.setUserAgent('Mozilla/5.0 ...');

Common errors

Important notes

• Puppeteer is Node.js only; for Python use Playwright or pyppeteer
• headless: "new" uses the new headless mode (more like real Chrome)
• page.evaluate() runs JavaScript in the browser context, not Node.js
• Always await browser.close() to prevent zombie processes
• waitUntil: 'networkidle0' = no network requests for 500ms (good for SPAs)