[Overall Style] Cinematic realism using Jacques Tati-style spatial comedy and Buster Keaton-style deadpan rhythm. Features warm evening natural light, restrained performances, and clear causal relationships, creating absurd contrast through a wuxia character interpreting modern slang literally. Designed for Seedance 2.0 multi-modal reference and native audio-visual synchronization.

[Characters] Role A | Swordswoman: @Image 1, 25-30 year old East Asian female, oval face, natural fair skin, sharp brown eyes, long black hair, tall slender build. Wearing white boots, white embroidered silk hanfu, translucent layered sleeves, silver waist ornaments, jade hairpin.
Role B | Biker: @Image 2, 25-30 year old East Asian female, expressive face, short brown ponytail, yellow jacket, blue jeans, white sneakers, same bicycle as reference.

[Shot 1 | 0-5s | Low Angle Wide Establishing] 16:9 landscape. Riverside plaza with a slight slope near a convenience store. Wet stone ground, concrete planters, glass shop reflections, vending machine, trees, distant city traffic, and a visible curb. Role B parks the bike on the slope without locking it; she points to the vending machine and says: "Sister, watch the bike for me." Swordswoman nods seriously: "Rest assured."

[Shot 2 | 5-10s | Medium Side Tracking] Role B walks a few steps toward the vending machine. The bike starts sliding down the slope. The swordswoman remains motionless, hands behind back, tracking the bike only with her eyes and head, intensely focused on "watching it" as if on a mission. Heroic music fades, leaving only wheel sounds. B looks back, sees the bike sliding, and freezes.

[Shot 3 | 10-15s | Close-up with Focus Shift] Close-up of Role B shouting: "Why didn't you stop it?" Focus shifts to Swordswoman who calmly replies: "You only told me to watch it, not stop it." She raises two fingers; a silver sword shoots into the frame, stopping the bike exactly one centimeter from the curb. Extreme close-up of her calm eyes as she adds: "Be clear next time." B stares speechlessly, the bike bell dings once, freeze frame.

[Technical] 15 seconds, three shots, 16:9 aspect, consistent motion direction, stable faces/clothing, realistic bike and fabric physics, accurate Mandarin lip-sync, synchronized environmental sounds.