agent-browser - Browser automation CLI for AI agents
Automate web interactions including navigation, form filling, screenshots, and data extraction
Tags
Updated: 2026-03-25Capabilities
Typical Inputs
Typical Outputs
What this skill does
- open website
- capture snapshot
- click element
- fill form
- type text
- select option
- check checkbox
- press key
- scroll page
- extract text
- retrieve URL
- retrieve title
- wait for element
- wait for load
- capture screenshot
- save PDF
- record session
- save state
- load state
Inputs
- URL
- form data
- user credentials
- CSS selectors
- text content
- session state files
- device identifiers
Outputs
- screenshots
- PDF files
- extracted text
- page URLs
- page titles
- video recordings
- session state files
Requirements
- macOS with Xcode for iOS support
- Appium with xcuitest driver