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.
automation ocr tesseract ui-automation coordinates autonomous-agents florence windowsui paddleocr screen-parsing omniparser
-
Updated
Jul 28, 2026 - Python