vision-llm
Eyes for AI coding agents 👁️ — render → perceive → report → fix → re-render. A machine-graded visual feedback loop (DOM/contrast/OCR-grounded + optional vision LLM) agents consume to self-correct before claiming done.
A lightweight, unified OCR toolkit with a one-liner API. Supports Surya, EasyOCR, PaddleOCR, Tesseract, and Vision LLMs through a single interface.
PyMidscene - Midscene.js 的 Python SDK 实现 | AI 驱动的自然语言 UI 自动化，告别选择器，用中文描述即可操作。与官方缓存格式完全兼容。
Audit PEA-eligibility of ETF KID documents with a vision LLM. French PEA (Plan d'Épargne en Actions) rules built in.
