How to Transfer Files with a Web Agent (without Nuking Your Context Trying)
Just as with airlines, we have luggage weight limits (and conditions). Instead of having the LLM hold file data, we give it a reference pointing to the data.
Just as with airlines, we have luggage weight limits (and conditions). Instead of having the LLM hold file data, we give it a reference pointing to the data.
Getting a file into or out of a remote browser is an infrastructure problem you have to think about. We've built the solution right into our CDP API.
Five hard-won lessons from building Browserless's agentic browser: why hostability, token efficiency, skills, mature protocols, and a stubborn model decide whether a browsing agent actually works.
How to use Scrapy to scrape dynamic, javascript-heavy sites, without using a library such as Playwright or Selenium, or having to host Chrome.
We'll show you how to automate a WebRTC video stream with Puppeteer & Browserless.
We'll share session management tactics, such as reusing cookies, reconnecting to browsers and using the --user-data-dir
[April Fools Prank] Introducing Headless Internet Explorer