agent-browser

CLI browser automation lets you navigate pages, fill forms, and collect on-screen data using repeatable commands. You open sites, inspect interactive elements via accessibility snapshots, and execute clicks or text inputs.

Screenshot capture is supported directly within web scraping workflows. The skill captures visual page states during navigation or after form submission, providing images for debugging, verification, and audit handoffs.

Browser automation handles login flows by filling inputs, clicking buttons, and selecting options through CLI commands. It uses ref-based accessibility snapshots to reliably target elements and wait for page responses during submission.

The best way to extract page content is using CLI commands to inspect text and query element attributes. This data collection method retrieves specific on-screen information reliably for scraping or verification tasks.

Parallel session management is supported for command-line automation. You can manage multiple browser sessions concurrently, allowing simultaneous website navigation and task execution across different automation workflows.

Accessibility snapshots provide ref-based element targeting that ensures reliable clicks, fills, and waits. This approach maps interactive controls predictably, reducing fragility compared to standard DOM parsing during page navigation.