Show HN: BrowseBrawl – What if browser agents battled to generate training data?

12•HrubyOnRails•1h ago

I remember watching the AlphaGo documentary in 2017. What stood out to me was that the model got drastically better when it started competing against itself. GANs clicked for me similarly: a generator and discriminator competing, and somehow the competition is what produces something remarkable.

I've been curious whether this principle generalizes to today's agents.

So mehulkalia and I built Browser Brawl at the YC / BrowserUse hackathon last weekend and won first place. It is a fun experiment in which an attacker agent tries to complete tasks on live websites while a defender agent injects JavaScript to sabotage it.

The analogy isn't perfect, because browser tasks aren't zero-sum. But our hypothesis is that an agent faced with an adversary should produce more interesting training data than one navigating clean, static environments.

Try it on: http://browser-brawl.com

GitHub: https://github.com/RichardHruby/browser-brawl

Demo Video: https://youtu.be/NIoFXv-JvBY

(Skip to [0:55](https://www.youtube.com/watch?v=NIoFXv-JvBY&t=55s) to see the agents “brawling” in the arena :), [1:52](https://www.youtube.com/watch?v=NIoFXv-JvBY&t=1m52s) to see the browser traces generated)

Would love to chat with anyone building or training browser agents. Happy to dive in below!

Comments

SobjectiveTruth•1h ago

This is hilarious

julian2k•1h ago

love it

mehulkalia•1h ago

Thanks!

mehulkalia•1h ago

Mehul here. One thing that surprised me while building this was how creative the defender agent became. It runs Claude Haiku on a timer and can choose from prebuilt disruptions like fake “Session Expired” popups, or generate custom JavaScript injections based on what the attacker is doing, like inserting fake “Search disabled” buttons. Digging through the traces and seeing the before/after screenshots of what the defender agent came up with was pretty funny, and kind of mind-blowing.

etash•50m ago

This is sick. Love the gamified creativity for the project + the actual real world usecase

sohilbhatia85•50m ago

Super cool!!

rohoswagger1•22m ago

This is so sick! Another crazy extension could be to give them certain handicaps / abilities to add more noise

The Disappearing American Mortgage

macOS code injection for fun and no profit (2024)

Ask HN: Maintainers, do LLM-only users often clutter your issues/PRs?

Show HN: Marra AI – AI consulting built for healthcare and occupational medicine

Computer Is Fast Enough

Models have some pretty funny attractor states

Pancreatic tumors cells 'choose' to either grow or tolerate treatment

Anthropic investors push to de-escalate Pentagon clash over AI safeguards

Zembed-1: The Best Text-Embedding Model

Agent memory is structured not fuzzy.why are we all using vector DBs for it?

Show HN: Health optimization as agent-guiding gradient descent

Why Write in the Time of LLMs?

From Claiming Land to Claiming Agents

Don't Fear AI: Why Humanity Will Thrive in the Age of Artificial Intelligence

Six Things I Learned Watching a Robotics Startup Die from the Inside

Show HN: SuperKey – JMeter Command Center

Show HN: Secure Agent Starter – A minimal template for building safer AI agents

How the War Department Learned to Stop Worrying and Love AI (With Naomi Klein)

Show HN: Turn .cursorrules / repo guidelines into GitHub pre-merge checks (OSS)

Ordered Dithering with Arbitrary or Irregular Colour Palettes

Teleprinter

What I talk about when I talk about prompting

Show HN: Novum – Automated ML Research Pipeline with Anti-Fabrication Guards

New York could prohibit chatbot medical, legal, engineering advice

Show HN:JSON API for US HTS and Canadian Customs Tariff, with Change Detection

Judge Blocks Virginia 1-Hour Social Media Law for Minors

Show HN: I spent 1000 hours building my dream personal finance dashboard

Higher risk of later-life memory,mental health issues in Amer. football players

FDA sends warning to 30 telehealth companies selling 'illegal' GLP-1s

Matthew Explains