Kling prompt for the Ramalama trend: copy it and go

Paste-ready paragraph and negative prompt

A vertical 9:16 video, 10 seconds long, one continuous shot, photoreal. A quiet residential road after dark with black asphalt, warm orange street lamps and trees and hedges on both sides. The camera is a phone held just above the ground, gliding backwards while the person from the reference photo walks toward it dressed as a classic witch: pointed black hat, black dress with a ragged hem, black-and-purple striped stockings and pointed black ankle boots. For the first six and a half seconds we see only boots, striped legs and the swinging hem, one confident runway step on every beat, and the face never enters the frame. Then the camera tilts fast from the shoes up to the face and shows the whole costume. The walk goes on with arms swinging, both hands cup the face and smile, and the last second and a half is a lean toward the lens and a big laugh. Face, hair and skin tone stay exactly as in the photo. Sound: soft instrumental electro-pop with a stomping kick on each step. No singing, no lyrics, no speech, and no text or watermark on screen.
Negative prompt: singing, lyrics, speech, text, captions, subtitles, logos, watermark, extra people, blurry face, hat falling off

Using Kling's two boxes

• Put the paragraph in the main prompt box and the Negative prompt line in the box underneath, because Kling reads the two separately.
• Choose the ten-second option. The 5-second one leaves no room for the long boots-only walk before the reveal.
• If a face shows up before the tilt, add one sentence: the camera stays below knee height until the tilt up.