Kling prompt for the Smalltown Boy AI dance: paste-ready, no song names

Prompt to paste

Mode: image-to-video
Format: vertical 9:16, roughly 15 seconds. Handheld wide-angle phone look, full frame, no fisheye circle, no black bars.
Cast: image 1 = LEAD, image 2 = PARTNER. Both keep their own face, fur or hair, colours, build and clothes. No swapping, no merging, no third character. Animals stay animals: upright on hind legs, front paws clamped on the sides of the head, mouth open in one long, theatrical wail, like a person in despair.
Place: a sunny backyard with a bright red painted wooden fence, green grass and blue sky. Phone fixed at chest height, close to the one yelling, whose head and shoulders fill the foreground. Nothing readable, no logos.
Shot 1 (0-7 s): the LEAD is right at the lens, both hands gripping their head, yelling nonstop, rocking in despair. Near 4 s both hands reach toward the camera, then return to the head. Small in the back, the PARTNER dances at full tilt, whole body visible.
Shot 2 (7-14 s, hard cut): positions trade. Now the PARTNER is at the lens, hands on head, yelling nonstop, and the LEAD dances in the back.
Shot 3 (14-15 s, hard cut): wide shot, the two standing shoulder to shoulder and dancing.
Sound: retro 1980s synth-pop instrumental, driving beat, with the characters' own long comic wails over it (cat yowl, dog howl, human "aaaah"). No lyrics, no speech.
Style: photoreal. Nothing written on screen, no logos, no watermark.

Tips

• Image 1 slot: the Lead portrait. Image 2 slot: the Partner portrait. A tool that takes a single image? Use one photo with both, and rename the roles in the opening lines.
• Leave titles and artist names out. Tools filter them, and borrowed music can bring rights trouble, so the prompt asks for an original-sounding instrumental. Put the trending sound on inside the app when you post.
• Tools change fast. Read the clip length and aspect ratio your plan allows, then edit the first line.