Browserless

@browserless can you grab the HTML from https://example.com/pricing and pull out all the plan names and prices using CSS selectors?

I've scraped the pricing page and extracted the following plans: • Starter Plan: $29/month • Professional Plan: $79/month • Enterprise Plan: Custom pricing The tool navigated to the page, waited for JavaScript to execute, then used CSS selectors to target `.plan-name` and `.plan-price` elements. All three pricing tiers were successfully extracted from the rendered DOM.

This example shows Browserless handling JavaScript-rendered content that a simple HTTP fetch would miss. The scraping tool waits for full page load before applying selectors, making it ideal for modern SPAs. You'll need to inspect the target page's DOM to write effective selectors — generic queries may return empty results if the site's markup doesn't match your assumptions.

@browserless generate a full-page PDF of our staging dashboard at https://staging.acme.io/dashboard so I can review the layout before we ship.

PDF generated successfully. The file captures the entire dashboard page, including all scrollable content below the fold. The PDF is formatted as Letter size and waited for the 'networkidle2' event before rendering, ensuring all charts and lazy-loaded widgets were fully displayed. The generated PDF is now available for download and shows the complete dashboard layout as it would appear to an authenticated user.

This demonstrates Browserless's ability to render authenticated pages if you pass session cookies or tokens in the request headers. The 'waitUntil' parameter is critical — choosing 'networkidle2' ensures dynamic content finishes loading, but adds latency. For pages with continuous polling or live updates, you may need to use a fixed delay instead to avoid timeouts.

@browserless run a custom Puppeteer script to log into https://vendor-portal.example.com, navigate to the invoices section, and download the latest PDF invoice.

Custom Puppeteer script executed successfully. The automation: 1. Navigated to the login page and filled credentials 2. Clicked through to the invoices section 3. Triggered the download of 'Invoice_2024_03.pdf' 4. Retrieved the file from the ephemeral download directory The downloaded invoice is now available. The script ran in a fresh browser context with a temporary download folder that was cleaned up after execution.

This showcases Browserless's most flexible tool — arbitrary Puppeteer code execution. You can automate complex multi-step workflows that require interaction (clicks, form fills, waits). However, you're writing real browser automation code, so you need to handle errors, timeouts, and selector changes yourself. The download directory is ephemeral, so retrieve files immediately after the script completes.

When Browserless beats manual PDF downloads for finance ops

A 6-person finance team at a SaaS company needs to archive vendor invoices from 40+ supplier portals every month. Half those portals require login and generate PDFs dynamically. Browserless's PDF generation tool handles the auth flow and waits for JavaScript to render before capturing the document. The team writes one Puppeteer script per portal, stores the API key in Switchy's shared workspace, and runs the batch monthly. This works until you hit portals with aggressive bot detection—then the Unblock tool adds cost per request. If you're archiving fewer than 100 PDFs monthly and the portals are stable, Browserless saves 4-6 hours of manual clicking. Beyond that threshold, evaluate whether a dedicated invoice aggregator is cheaper than per-request unblocking fees.

Screenshot and scrape tools for weekly competitor tracking

A 3-person product marketing team tracks 12 competitor landing pages for pricing changes and feature announcements. They use Browserless's screenshot tool to capture full-page images every Monday and the CSS scraper to pull headline text and CTA copy into a shared spreadsheet. The scraper waits for JavaScript to load, so it catches React-rendered content that basic curl requests miss. This setup costs about $15/month on Browserless's starter tier and runs in 90 seconds via a scheduled Switchy prompt. The trade-off: if competitors implement Cloudflare or reCAPTCHA, you'll need the Unblock tool, which doubles the per-request cost. For low-frequency monitoring where you need visual proof and structured data, Browserless is faster than maintaining your own Puppeteer instance.

When custom Puppeteer scripts help debug user-reported UI bugs

A 5-person support team at a B2B app gets 8-10 tickets weekly about layout bugs that only appear in specific browser states or after certain user actions. They use Browserless's custom function tool to write Puppeteer scripts that reproduce the exact click sequence, viewport size, and auth state the customer reported. The script runs in Browserless's cloud, captures a screenshot, and attaches it to the Linear ticket—all from a Switchy prompt the support rep triggers. This works because Browserless handles the browser binary and network isolation; the team doesn't provision infrastructure. The limit: if your app requires SSO or complex session cookies, injecting auth into the Puppeteer script gets brittle. For apps with token-based auth and reproducible UI states, this cuts bug triage time from 20 minutes to 3.