I added in the system prompt that it should skip CAPTCHAs and hand control back to the user. Currently working on a proper human-in-the-loop feature. That's actually one of the key advantages of running the agent inside your own browser.
For curiosity's sake, have you had it try to attempt captchas?
If so, what were the results?
I use a text-based approach. Captchas like “crossroad” usually need a screenshot, a visual model and coordinate-based mouse events.