agent-browser - Browser automation for AI agents
Automates browser interactions such as navigation, form completion, clicking, screenshots, data extraction, web app testing, and authentication.
Tags
Updated: 2026-10-07Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Navigate websites
- Inspect interactive elements
- Fill and submit forms
- Click and select controls
- Capture screenshots and PDFs
- Extract page data
- Manage browser sessions
- Compare page states
- Automate iOS Safari
Inputs
- Website URLs
- Form data
- Authentication credentials
- Saved authentication state
- CSS selectors
- Local files
Outputs
- Navigated browser pages
- Submitted forms and browser actions
- Extracted text or JSON
- Screenshots and PDF files
- Downloaded files
- Saved session state
- Page state differences
Requirements
- agent-browser CLI
- Bash access to agent-browser
- macOS, Xcode, and Appium for iOS