Browser Automation

Browser Automation

All agent containers come with agent-browser pre-installed and configured to use system Chromium. The native Rust CLI is built from source during the Docker build for musl/Alpine compatibility. Agents can navigate websites, fill forms, take screenshots, and extract data — no setup needed.

The browser session persists across messages within the same agent session and shuts down with the container on idle timeout.

@general open https://example.com and tell me what you see
@general go to https://example.com/form, fill in the email field, and submit

The agent-browser skill is automatically loaded into every agent's system prompt, teaching it the full command set (open, snapshot, click, fill, screenshot, sessions, etc.).