Girl Caught Hiding Candy AI Image Prompt

Basic parameters: duration 10 seconds, aspect ratio 9:16 vertical, cinematic 1080P HD 60fps, style warm and healing, cute childlike interest, real life documentary, first-person perspective with slight steady follow shot no shake, audio retain all action voices and body sounds no subtitles no background music. Character: female lead from image 1, strictly generate video according to the character. Scene: daytime indoor, bright living room with sofa and simple home furnishings. Plot shots + action lines script: 1. Shot: female lead sits on sofa corner, secretly stuffs candy into pocket with sneaky movements. Me (serious female voice): Hiding candy again, let me see. 2. Shot: female lead dodges slightly, eyes wandering, covers pocket with small hands, guiltily argues: I'm not hiding, pocket is empty. 3. Shot: eye contact pauses, female lead loses confidence, slowly releases hand from pocket. Me: Don't lie, I saw everything clearly. 4. Shot: female lead frowns reluctantly, slowly pulls out colorful candy from pocket and holds it in palm. Female lead: Just this one, I just want a little treat. 5. Shot: female lead reluctantly places candy on sofa, pouts mouth, lowers head. 6. Closing shot: female lead leans sideways on sofa, dejected and drooping head, showing mood, looking wronged and cute. Visual requirements: smooth and natural, facial expressions match plot changes, actions coherent and not stiff, overall atmosphere soft and homey.