30-Second AI Rescue Shorts — Realistic Story Workflow

Every concept must feel like a real-life animal rescue incident captured by a third person with a camera.

The video must NOT feel like a movie scene, staged commercial, animation, cartoon, fantasy sequence, or cinematic action film.

The visual feeling should be: Photorealistic, raw and emotionally believable, natural animal behavior, real-world locations, realistic environmental conditions, realistic human rescue behavior, natural lighting, authentic handheld/observational recording feel.

Concept Diversity: Make every idea substantially different from the others. Vary: Animal species, baby animals and adult animals, mother-and-baby situations, animal trapped in objects, animal trapped in natural environments, animals stuck in mud, water, ice, snow, rocks, fences, pipes, nets, debris, holes, abandoned structures, etc. Different rescue methods, different locations, different weather conditions, different times of day, different emotional situations, different levels of urgency and difficulty. Do not repeatedly use the same animal, location, trap, or rescue mechanism.

Emotional Story Types: Include a mixture of situations such as: A mother struggling to save her trapped baby, a baby animal separated from its mother, a baby animal trapped somewhere dangerous while the mother stays nearby, an adult animal accidentally trapped in a human-made object, an injured or exhausted animal unable to escape, an animal trapped by natural conditions, a mother refusing to leave her baby during the rescue, a frightened animal initially resisting the rescuer, a rescue where the animal appears helpless and the situation gradually becomes hopeful.

Human Rescuer Rule: Every concept must include a human rescuer who physically enters the situation and attempts to rescue the animal. The rescue must feel practical and believable for the specific situation.

Camera Rule: The final video concept must be designed as if: One person is recording the rescue from a third-person observational position while another person is physically performing the rescue. The camera should remain observational and realistic rather than behaving like a professional movie camera. Do NOT describe dramatic cinematic camera movements, impossible camera positions, aerial shots, or multiple-camera setups.

Output Format: For each idea, provide: Idea 1 — [Short Emotional Title]. Animal:, Location:, Situation:, Rescue:, Emotional Hook:. Keep each concept concise but visually clear enough that it can later be converted into a complete video-generation prompt.

Selection Rule: After generating the ideas, tell the user: "Select the idea(s) you want to generate. You can select one or multiple ideas. If you select multiple ideas, submit all selected idea numbers together. After your selection, proceed to Phase 2." Do not generate the full video prompt in Phase 1. Do not generate character sheets, storyboards, dialogue, or detailed shot-by-shot instructions in Phase 1.