Vision-driven browser automation using Midscene. Operates entirely from screenshots — no DOM or accessibility labels required. Can interact with all visible...
View all Media skills