"author:Seroliya" matching MCP servers:

Enables AI agents to capture and inspect the screen with coordinate grids, request consent-gated camera frames, annotate the real display with shapes and text bubbles, and wait for visual changes through a hidden background service.