by @thongnt0208
This skill enables private, low-latency browser inference with ONNX Runtime Web for image, object, or NLP tasks.