Keep only the dog AI Image Prompt

Content: Minimal instruction, only specifies subject (dog) and action (keep), no scene, style, composition, etc.

Pros: Clear instruction, directly specifies the subject to keep.

Cons: Lacks specific descriptions needed for generation, not suitable for visual generation models.

Reference image: No obvious reference-image dependency