CPU-friendly screen-to-JSON parser for AI desktop agents. Extracts pixel-perfect coordinates, OCR, UI semantics, cursor context, and structured screen state for reliable LLM-driven automation.
automationocrtesseractui-automationcoordinatesautonomous-agentsflorencewindowsuipaddleocrscreen-parsingomniparser
-
Updated
Jul 28, 2026 - Python