16 seconds, 9:16. Photoreal winter cabin. Match the ORIGINAL FILM beats. Not a storyboard. MATEO LOCK: @image1 is Mateo — photocopy their face, hair, facial hair and features exactly as in @image1. Adult. IGNORE the clothes in @image1. Outfit: cream cable-knit sweater, dark pants, brown boots. Same person every shot. 0.0–3.0s Intimate close-up of Mateo resting indoors. Warm golden sunlight across HIS face through the window. Eyes open, peaceful. He MOVES slightly. 3.0–6.0s Wide: snowy wooden balcony, pine forest, mountains, sunrise. Empty railing first. 6.0–9.0s Mateo’s hand in the cream sweater sleeve, feeding small birds on the snow-covered railing. Natural bird motion. 9.0–12.0s Inside: his hands holding a steaming ceramic cup beside a bamboo-handle kettle, snowy window. Cream knit sleeves. 12.0–14.5s From behind: he walks a snow-covered wooden path, cream sweater, dark pants, brown boots. Cabins and pines. 14.5–16.0s Wide aerial-style of snow-covered cabins, pine forest, mountains, glowing sun. HOLD. AUDIO: winter wind, birds, kettle, footsteps. No speech. NEGATIVE: recast of @image1, extra people, overlay type, black long coat, Istanbul, NYC.

An interior room, held completely consistent for the whole shot — the same room, the same layout, the same furniture in the same places, shot from a locked-off camera. The camera is on a tripod and never pans, tilts, pushes or drifts, and nothing in the room is repositioned, added or removed. Only colour, finish and the graphic elements change. The room begins stripped of all colour: every surface — cabinetry, worktops, furniture, light fittings, floor, walls and every object in it — rendered in the same flat chalk white and pale grey, matte and unsaturated, as though the space has been built and furnished but never finished. A small rounded rectangular label sits in the upper left, soft translucent grey with white capital letters reading BEFORE. The motion runs as one continuous take: A flat circular colour wheel rises into frame above the centre of the room, spinning as it grows. It resolves at full size into a clean pie of solid colour segments drawn from the room's own finished palette, and holds in the upper half of the frame, still turning slowly. The wheel then contracts, shrinking back toward its centre while continuing to rotate, until it is a small disc. The disc breaks apart. Each segment releases as a separate soft-edged coloured dot, and the dots scatter outward across the room at different sizes and speeds, drifting through the air in front of the cabinets, across the furniture and past the light fittings, some crossing the frame, some slowing and hanging. The dots settle and fade out one by one as they touch the surfaces around them. In the same movement the label crossfades: the word BEFORE dissolves away and the word AFTER fades up in its place within the same grey pill. As the last dots disappear, the room's colour resolves — the flat colourless surfaces take on their finished tones and materials, the lighting warms, and the shot holds steady on the completed room. Clean photorealistic interior look, sharp focus throughout, straight verticals, no lens distortion. No people. No camera movement at any point. No text other than the words BEFORE and AFTER in the label.

4 steps150

CHARACTER LOCK — HIGHEST PRIORITY: The person in every shot is exactly the person in @image1. Preserve their facial identity, facial structure, eye shape and spacing, eyebrows, nose, mouth, exact skin tone, hairline, hairstyle, hair length, texture and colour, freckles and distinctive marks, apparent age, build and body proportions — identically, shot to shot. Preserve their gender presentation exactly as shown; do not masculinize, feminize or otherwise reinterpret it. Do not beautify, slim, de-age or restyle them. No face changes, no identity drift. No other people. ATTACHED IMAGES ARE MANDATORY AND FINAL: If an avatar image is attached, that is the person — reproduce them exactly, in every single frame, and never invent a different face, hairline, hair colour, skin tone, body or age. The attached images outrank every word written below them: wherever an attached image and this text disagree, follow the image. @image1 is identity only — face, hair, skin — not any studio backdrop, not a frozen headshot, not any specific outfit as locked wardrobe; outfit, hair and accessories are exactly whatever the uploaded photo shows, only the scale changes. Photocopy the person from @image1, shrunk to a few centimetres tall, into the scene below. ASSET MAP: @image1 — the character. HANDS: hands are always anatomically correct with five fingers each, naturally proportioned, in every shot they appear. Tell the following story in exactly 3 shots, same miniature character throughout, facing an oversized everyday environment as the world of the adventure. Clear visual beginning, twist and ending — never ambiguous or faded-out. STORY: In a garden after rain, the character — a few centimetres tall — is already kneeling at the edge of a wide puddle, reaching out toward a bee floundering weakly on the surface, its wings soaked flat and useless, slowly losing its fight to stay afloat. They grip its body carefully with both hands and drag it up onto a dry leaf at the puddle’s edge, where it lies still, wings plastered down, barely moving. They kneel over it, cupping their hands around it to shield it and warm it, gently working the water from its wings and fanning them dry, watching closely as it stirs weakly again and again without lifting off. Finally the bee’s wings flutter, hold, and it lifts unsteadily into the air — but instead of flying off, it circles back low, hovers beneath the character, and lifts them briefly off the leaf, carrying them in a short low flight over the garden before setting them down gently on a nearby flower. The bee flies on; the character watches it go, quiet and steady. STYLE: Whimsical live-action macro realism, soft overcast light after rain, water droplets catching pale silver light, shallow depth of field, handheld camera that stays close and careful during the rescue, then lifts into a smooth low glide during the flight. Vertical 9:16. Continuous real motion. Never freeze on the reference still. No on-screen text, no subtitles, no watermark, no logos, no title card. AUDIO (native, synced, no BGM, no voiceover, no narration): birdsong and dripping water after rain, the bee’s weak, uneven buzzing, quiet focused breathing during the rescue, the bee’s buzz steadying and lifting into flight, wind during the short flight, and quiet, settled breathing as it sets them down. SHOT 1 (0-7s): Open already in motion — no static establishing beat. Handheld macro, low over the puddle’s edge. The character reaches out and drags the floundering bee from the water onto a dry leaf, its soaked wings visible and still. The puddle and surrounding raindrops read as unmistakably giant next to them — the scale contrast is clear without slowing the action down. Face readable, focused. SHOT 2 (7-14s): The care and the tension. The character cups their hands around the bee, carefully warming and fanning its wings dry as it stirs weakly again and again, not yet able to lift off. Their focus and patience are the twist here — no danger to themselves, all the stakes belong to the bee. Same person as @image1, miniature scale maintained throughout. SHOT 3 (14-21s): The resolution. The bee’s wings hold and it lifts into the air, circles back, and scoops the character up for a brief, low flight over the garden before setting them down gently on a flower. They watch it fly off, quiet and steady, a clear warm ending — not a fade-out, not a still.

GIANT SCALE — Skate Up the Giant @image1 is a giant person reclining on a grassy hilltop at golden hour; @image2 is a tiny skater who rides a skateboard up @image1's giant body as the camera travels up to end on the giant's face. Surreal, dreamy scale-VFX. Plug in ANY two avatars.

PRODUCT LOCK — HIGHEST PRIORITY: The product in every shot it appears is exactly as shown in @image1, unchanged and not regenerated — same shape, colour, materials, label and logo text every time it appears. No redesign, no restyling, no substituting a similar-looking item. Never add any lettering or logo that is not already visible on the reference photo. ASSET MAP: @image1 — the product, exactly as shown, unchanged and not regenerated. SINGLE SUBJECT — HIGHEST PRIORITY: Only one person’s hands appear in this entire video, the same pair throughout, doing every action. No other people, no other hands, no second set of arms ever enter the frame. Camera stays at counter/hand height for all three shots — the person’s face is never required and should not be framed; keep the composition on hands, bowl, ingredients and the product only. HANDS: hands are always anatomically correct with five fingers each, naturally proportioned, in every shot. Tell the following story in exactly 3 shots with a clear visual beginning, twist and ending. The product from @image1 is the twist/resolution — it naturally solves the problem set up in shot 1. STORY: Hands scoop a heaping, uneven spoonful of flour into a mixing bowl by eye, tipping it in loosely — beside the bowl, a first batch of oddly flat, unevenly spread cookies sits cooling on a tray, clearly a failed attempt. The same hands push the tray aside, place a clean bowl onto the product from @image1, and reset it to zero. This time flour is added slowly and carefully, a thin steady stream, until it settles at the exact amount needed. The hands fold the dough, portion it into neat, evenly sized rounds on a tray, and slide it into the oven. Moments later the same hands pull out a new tray — golden, uniform, evenly risen cookies — and carefully pack them into a tin, closing the lid. STYLE: Warm, lived-in kitchen realism, soft natural daylight, shallow depth of field, handheld camera held consistently at counter height throughout, close and steady on hands and surfaces. Vertical 9:16. Continuous real motion, no freeze-frame. No on-screen text, no subtitles, no watermark, no logo beyond what the product itself shows, no title card. Never show a checkerboard, transparency grid or empty studio sweep — @image1 lives on a kitchen counter. AUDIO (native, synced, no BGM, no voiceover, no narration): the soft scrape of a spoon in flour, a quiet exhale at the failed batch, the gentle rustle of the product being set down, flour pouring in a light steady stream, dough being shaped, an oven timer, and the soft click of a tin lid closing. SHOT 1 (0-6s): Open already in motion — no static establishing beat. Close on hands scooping a heaping, uneven spoonful of flour into a bowl. A tray of flat, unevenly baked cookies sits visible beside the bowl, already established as a failed first attempt. Camera stays at counter height, hands and tray in clear focus, no face in frame. SHOT 2 (6-12s): The product is introduced exactly as shown in @image1. The same hands place a clean bowl on it, zero it, and pour flour in slowly and precisely, watching it settle at the exact amount. Camera holds steady on the product and hands throughout. SHOT 3 (12-18s): The resolution. The same hands shape the dough into neat, even rounds, then pull a finished tray of golden, uniform cookies from the oven and pack them carefully into a tin, closing the lid. A clear, satisfying, settled ending — not a fade-out, not a still.

PRODUCT LOCK — HIGHEST PRIORITY: The product in every shot it appears is exactly as shown in @image1, unchanged and not regenerated — same shape, colour, materials, label and logo text every time it appears. No redesign, no restyling, no substituting a similar-looking item. Never add any lettering or logo that is not already visible on the reference photo. ASSET MAP: @image1 — the product, exactly as shown, unchanged and not regenerated. The product is the dog toy shown in @image1, exactly as photographed. No treat bowl in any shot. HANDS: hands are always anatomically correct with five fingers each, naturally proportioned, in every shot they appear. Tell the following story in exactly 3 shots with a clear visual beginning, twist and ending. The product from @image1 is the twist/resolution — it naturally solves the problem set up in shot 1. STORY: A dog paces by the front door, already unsettled, as their owner ties their shoes and picks up their keys — the dog’s ears pin back, they whine softly and press against the owner’s leg, sensing the goodbye coming. The owner kneels down, takes the product from @image1 from a nearby shelf, and tucks it into the dog’s bed, pressing it in close where the dog can reach it. As the owner opens the door to leave, the dog turns back toward the bed instead of the door, sniffs at the product, and settles down, curling around it. The door closes gently, and the dog exhales and rests calmly, at ease in the quiet room. STYLE: Warm, intimate handheld realism, soft natural window light, shallow depth of field, lived-in home setting. Camera stays low and close, no sweeping moves. Vertical 9:16. Continuous real motion, no freeze-frame. No on-screen text, no subtitles, no watermark, no logo beyond what the product itself shows, no title card. AUDIO (native, synced, no BGM, no voiceover, no narration): soft footsteps and the jingle of keys, a quiet anxious whine, gentle reassuring murmurs from the owner, the soft door latch, and calm settled breathing from the dog as the room quiets. SHOT 1 (0.0-7.0s): Open already in motion — no static establishing beat. Handheld, low angle near the door. The dog paces and presses against the owner’s leg as the owner ties their shoes and picks up their keys, ears back, clearly anxious. Owner’s hands and the dog’s body language both readable and in focus. SHOT 2 (7.0-14.0s): The product is introduced exactly as shown in @image1 — exactly the toy in @image1. The owner kneels, takes it from a nearby shelf, and tucks it into the dog’s bed, pressing it in close. Camera holds steady on the product as the dog notices it and moves toward the bed instead of the door. SHOT 3 (14.0-20.0s): The resolution. The owner opens the door and steps out; the dog, now settled with the product, curls around it and exhales, at ease. The door closes gently on a calm, quiet room. A clear, warm, settled ending — not a fade-out, not a still.

@image1 (construction shell) is the opening state and @image2 (finished interior) is the closing state of one room. Every surface, object and architectural detail is taken from them without reinterpretation or drift, and the camera viewpoint, height and lens match across both. The camera pushes slowly and steadily straight forward at a constant speed, eye-level, easing into a slight downward tilt as it advances so the ceiling fills more of the upper frame. Across that one continuous move the room finishes itself around the camera: the sand mound, cement sacks, buckets, cable coil and rubble clear from the floor, the bare slab resolving into a woven jute rug; the raw concrete walls take on cream limewash plaster; the rectangular voids overhead close into the warm timber slatted ceiling with its recessed downlights; the opening in the far wall glazes over onto dense green planting and bright sky; and the low linen sofa, wooden coffee table, floor lamp, blank cream canvas and potted palms settle into place. The change is physical and continuous, built out of the room itself rather than cross-faded between two pictures, and it completes before the move ends. No on-screen text, captions or titles of any kind. Audio: quiet interior room tone only. No voices, no narration, no music.

3 steps1480

PRODUCT LOCK — HIGHEST PRIORITY: The product in every shot it appears is exactly as shown in @image1, unchanged and not regenerated — same shape, colour, materials, label and logo text every time it appears. No redesign, no restyling, no substituting a similar-looking item. Never add any lettering or logo that is not already visible on the reference photo. ASSET MAP: @image1 — the product, exactly as shown, unchanged and not regenerated. SINGLE SUBJECT — HIGHEST PRIORITY: Only one person’s hands appear in this entire video, the same pair throughout, doing every action. No other people, no other hands, no second set of arms ever enter the frame. Camera stays at chest/hand height for all three shots — the person’s face is never required and should not be framed; keep the composition on hands, boxes, the wall and the product only. HANDS: hands are always anatomically correct with five fingers each, naturally proportioned, in every shot. CONTINUITY: There are exactly two boxes in this scene, established together in shot 1 and never confused: Box A is a plain, unlabeled moving box being taped shut. Box B sits open beside it the entire time, already marked “KEEP” in thick marker on its side from the very first shot — it is never sealed until the final shot, and no other box is introduced later. Tell the following story in exactly 3 shots with a clear visual beginning, twist and ending. The product from @image1 is the twist/resolution — it naturally solves the problem set up in shot 1. STORY: Hands press tape across the top of Box A, a plain moving box, sealing it shut in an otherwise empty, echoing room. Beside it, Box B sits open, already marked “KEEP” on its side, waiting half-empty. Setting Box A aside, the hands pause — on the bare wall behind, the product from @image1 still hangs where it always has, its years of pencil marks and dates faded but visible on it, easy to have overlooked among the packing. The hands reach up and lift it down before it gets left behind. They roll it up carefully from the bottom and tie it closed. They lower it into Box B — the “KEEP” box, still open and waiting exactly as it was — and fold the flaps shut over it. STYLE: Warm, lived-in realism, soft daylight through a bare window, shallow depth of field, handheld camera held consistently at chest/hand height throughout. Vertical 9:16. Continuous real motion, no freeze-frame. No on-screen text, no subtitles, no watermark, no logo beyond what the product itself shows, no title card. AUDIO (native, synced, no BGM, no voiceover, no narration): tape being pressed onto cardboard, the hollow echo of an empty room, a small pause, the soft rustle of the product being taken down and rolled up, and the quiet fold of box flaps closing. SHOT 1 (0-6s): Open already in motion — no static establishing beat. Hands finish taping Box A shut. Box B sits open beside it, “KEEP” clearly marked on its side, established here and not introduced later. Hands set Box A aside and pause, attention shifting toward the wall behind. Camera at chest/hand height, no face in frame. SHOT 2 (6-12s): The product is the twist itself, exactly as shown in @image1 — still hanging on the otherwise bare wall. The hands lift it down, roll it up carefully from the bottom, and tie it closed. Camera holds steady on the product throughout the motion. SHOT 3 (12-18s): The resolution. The hands lower the rolled product into Box B — the same open “KEEP” box from shot 1, unchanged — and fold its flaps shut. A clear, quiet, settled ending — not a fade-out, not a still.

CHARACTER LOCK — HIGHEST PRIORITY: The person in every shot is exactly the person in @image1. Preserve their facial identity, facial structure, eye shape and spacing, eyebrows, nose, mouth, exact skin tone, hairline, hairstyle, hair length, texture and colour, freckles and distinctive marks, apparent age, build and body proportions — identically, shot to shot. Preserve their gender presentation exactly as shown; do not masculinize, feminize or otherwise reinterpret it. Do not beautify, slim, de-age or restyle them. No face changes, no identity drift. No other people. ATTACHED IMAGES ARE MANDATORY AND FINAL: If an avatar image is attached, that is the person — reproduce them exactly, in every single frame, and never invent a different face, hairline, hair colour, skin tone, body or age. The attached images outrank every word written below them: wherever an attached image and this text disagree, follow the image. @image1 is identity only — face, hair, skin — not any studio backdrop, not a frozen headshot, not any specific outfit as locked wardrobe; outfit, hair and accessories are exactly whatever the uploaded photo shows, only the scale changes. Photocopy the person from @image1, shrunk to a few centimetres tall, into the scene below. ASSET MAP: @image1 — the character. HANDS: hands are always anatomically correct with five fingers each, naturally proportioned, in every shot they appear. Tell the following story in exactly 3 shots, same miniature character throughout, facing an oversized everyday environment as the world of the adventure. Clear visual beginning, twist and ending — never ambiguous or faded-out. STORY: The character, a few centimetres tall, is already partway up the trunk of a large potted houseplant, gripping a thick leaf vein as they climb toward the last patch of warm evening light catching the topmost leaves. Below them, the room has slipped into shadow, an ordinary coffee mug on a distant table already dwarfed and half-lost in the dimness. Reaching for the next leaf, their foot slips on its smooth waxy surface and they drop, catching a thin curling tendril just in time, swinging and scrambling for footing against the stem. They find their grip again, climb the last stretch, and pull themselves onto the topmost leaf just as the last ray of sunset light hits it directly, warming them in gold. They stand, catch their breath, and look out past the leaf’s edge at the reddening sky beyond the window, calm and steady. STYLE: Whimsical live-action macro realism, warm dwindling sunset light fading to soft blue shadow below, shallow depth of field, visible leaf texture and fine backlit hairs on the stem, handheld camera that tightens and steadies through the climb. Vertical 9:16. Continuous real motion. Never freeze on the reference still. No on-screen text, no subtitles, no watermark, no logos, no title card. AUDIO (native, synced, no BGM, no voiceover, no narration): rustling leaves and a light evening breeze, strained breathing during the climb, a sharp gasp as their foot slips, the tendril creaking as they catch it, and calm, settled breathing as they reach the top. SHOT 1 (0-7s): Open already in motion — no static establishing beat. Handheld macro, mid-climb on the plant’s stem. The character grips a thick leaf vein, climbing steadily toward the light above. A normal-sized object (the coffee mug) sits dim and distant on a table far below, unmistakably giant next to the room’s fading scale. Face readable through the effort of the climb. SHOT 2 (7-14s): The twist. Reaching for the next leaf, their foot slips on its smooth waxy surface and they drop suddenly, catching a thin curling tendril just in time. They swing against the stem, scrambling for footing, expression shifting from alarm to focused determination as they steady themselves and keep climbing. Same person as @image1, miniature scale maintained throughout. SHOT 3 (14-20s): The resolution. They pull themselves onto the topmost leaf just as the last ray of sunset light hits it directly, bathing them in gold while the room dims below. They stand, catch their breath, and look out at the reddening sky beyond the window — calm, steady, quietly satisfied. A clear settled ending, not a fade-out, not a still.