agent-browser - CLI browser automation tool for AI agents.
Automates browser navigation, element interaction via ref IDs, form filling, data scraping, and screenshot capture.
Tags
Updated: 2026-09-16Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Navigate to target web URLs
- Capture accessibility DOM snapshots
- Interact with page elements
- Take screenshots and PDF exports
- Manage parallel browser sessions
Inputs
- Target web URLs
- Element ref identifiers
- Input text values
Outputs
- Accessibility snapshot data
- Saved PNG screenshot files
- Exported PDF document files
- Extracted element text and attributes
Requirements
- Node.js runtime
- agent-browser CLI package
- Chromium browser installation