fortress

i use claude code daily and measured what pages cost it while doing research. an average wikipedia article, for instance, is 68,240 tokens of raw html (tiktoken); nike's homepage is 353,000. claude code's built-in webfetch handles the easy case well. it summarizes wikipedia to about 950 tokens and clears cloudflare on some sites like indeed and ticketmaster. but, and there's always a but, on js-rendered and some anti-bot pages it returns nothing. quotes.toscrape.com/js gives "no quotes found"; nike.com gives a 403. your agent then dumps the raw html back into context and still fails. (note: i have also had cases where i read through the chat at the end and saw that it failed and just pulled from either training data or stale caches from other sources) so i worked on building an open-source stealth browser (recompiled chromium) that runs a

i personally believe that the internet is the best thing we've ever built, and arguably, nothing we made before comes close. now agents can use it too, but the web treats them as second-class. every agent works for a person, so whatever it keeps from agents it keeps from us. we believe that if we let them in, each of us gets the whole of what we built, a thousand times over. and to add to that, democratization of tools that allow this is essential, despite the fact that there will always be bad actors. so, we built fortress, which is an open-source stealth chromium engine. it is a recompiled fork that corrects the browser fingerprint from inside the engine: the surfaces bot detectors read, canvas, webgl, audio, fonts, navigator, and about thirty more, are patched in chromium's c++, with no javascript layer sitting on top for a page to cat