User shares Minimax Ref2VA prompt for character identity consistency

The prompt uses three input images to maintain identity, face, and clothing consistency between two characters in a scene. It is designed for use with the Minimax Ref2VA model.