What problem does it solve?
This Skill enables automated browser control and multi-step web interactions by leveraging CDP and ARIA-based element references to perform navigation, interaction, observation, and verification while maintaining safety and isolation.
Core Features & Use Cases
- Observe: capture ARIA accessibility tree, element refs, and page state for stable interactions.
- Interact: click_ref, fill_ref, click, type, hover, scroll, drag, and file upload across complex UIs.
- Verify & Debug: screenshot, get_console, get_network, cookies_get, and annotate_som for visual validation and debugging.
- AI-assisted browsing: ai_browse for goal-directed, multi-step browsing with plan-act cycles.
- Use Case: automate login flows, form autofill, content extraction, and QA checks across modern web apps.
Quick Start
Start with observe to capture the page structure and ARIA references, then drive interactions using ARIA refs and screenshots.