Generate text-to-video and image-to-video prompts with shot timing, subject motion, camera movement, lensing, lighting, audio cues, continuity anchors, and model-aware constraints.
