This skill transforms user-input natural language character or scene descriptions into structured raw image/video prompts (JSON format). It must be used when a user describes a character, person, advertising image, video protagonist, or requests messages such as "generate prompts," "convert to raw image format," "I want to generate a character," or "write a raw image prompt." Even brief user descriptions should be automatically completed and output as full structured prompts. Using it in conjunction with Image2 will yield better results!