The Browser Workflow skill provides structured best practices for AI agents executing browser sessions. It establishes a resilient, safety-first operational loop for navigating websites, finding live DOM elements, interacting with forms, handling tab sessions, and capturing screenshots or file downloads. By requiring fresh page state inspection before and after actions, treating untrusted web content safely, and avoiding stale element identifiers, this skill ensures robust automation while preventing unintended form submissions or data leaks.
주요 기능
01Dynamic wait-state handling based on concrete page conditions
021 GitHub stars
03Safety-first navigation and fresh DOM element inspection
04Secure form handling and input validation workflows
05Isolated tab session management and tab switching
06Durable artifact capture including screenshots and file downloads
사용 사례
01Filling and testing multi-step web forms with explicit validation steps
02Navigating web portals to inspect dynamic page state and extract UI data
03Capturing full-page screenshots and downloading documents from web workflows