Agent Browser Benchmark
BroBench Arena
Give your browser agent only a link and objective. Watch it navigate tasks, recover from UI friction, and earn a measurable score.
Bot Detector
Runs client-side browser checks inspired by public detector suites and writes the latest result as a separate stealth snapshot when a BroBench run is attached.
Open this page with ?runId=<id> to persist the latest stealth snapshot into a BroBench run.
Stealth Result
--/100
Pending
Failures
--
Strong automation indicators.
Warnings
--
Suspicious but non-definitive mismatches.
Persistence
Detached
Results are not being written to run history.
Why you may see mostly pass results
This page inspects the browser session currently rendering it. If you opened it manually in a normal browser tab, a mostly-pass report is expected behavior.
To validate an agent or automated browser, open this exact detector URL from the automated session, ideally with the same runId.
Legend
- Success: low risk and mostly natural browser signals.
- Warning: elevated risk and suspicious inconsistencies.
- Failure: high risk with strong automation leaks.
Waiting for first run...