Kernel
Cloud browsers for agents to browse, fill forms and take screenshots.
Kernel, in a GBrain workspace
Cloud browsers for agents to browse, fill forms and take screenshots.
gbrain.io
Cloud browsers for agents to browse, fill forms and take screenshots.
Kernel runs Chromium browsers in the cloud for AI agents, with saved profiles, proxies and video replays of each session. Connected to a workspace, GBrain can answer “what does the vendor portal list for this invoice?” by sending a browser to look, fill in a web form that has no API, or capture a screenshot of a page as it is right now.
Signing in happens on Kernel’s own page, and browser time counts against the Kernel account.
GBrain gives Kernel a command line too. From a terminal, a
script or Claude Code,
gbrainio tool help kernel
lists its commands once it is connected.
You sign in on Kernel's own page. No key to copy. Then add it to a workspace, and anyone on the team can ask about it in plain words.