Our novel method improves visual prompting accuracy through affine/color transformations and TrivialAugment data augmentation, achieving state-of-the-art results with minimal overhead.