A LoRA trained to mimic the style of classic Capcom illustrations. Due to limitations of the training data, put "white borders" in the negative prompt unless you wish to have that in the final image.