Anime Sword-Fighting Cinematic

Raw tactical bodycam POV, pure real-world footage, chaotic handheld energy, natural daylight, fast imperfect movement, live and unpolished, continuous first-person operator view with rifle always visible in frame. 0-4s: [Bodycam POV] Helmet-cam on a heavy wooden garden gate. Operator’s gloved hands grip a metal battering ram with teammates. Hard coordinated strikes smash the lock and hinges. Door bursts inward in a cloud of splinters and dust. Immediate aggressive push forward through the breach, rifle raised, constant reactive camera shake and urgent team radio chatter. 4-10s: [Bodycam POV] Inside a real suburban garden. Classic ceramic garden gnomes — painted, bearded, pointed hats — suddenly animate and move with unnatural speed. They draw tiny but real weapons (knives, small axes, pistols) and attack. Operator fires controlled bursts, muzzle flash lighting the frame, spent casings eject. Gnomes take hits but keep coming, completely unstoppable, closing distance from all sides amid chaotic close-quarters combat, dirt, plants and flying debris. 10-15s: [Bodycam POV] One large garden gnome sprints straight at camera with impossible speed. It leaps, slams into the operator chest-first, drives the body backward and violently to the ground. Camera tumbles hard, sky and gnome face filling the lens as the operator is fully taken down. Rifle still visible, thrashing hands, raw impact physics. Photorealistic raw bodycam, natural physics, coherent motion, authentic tactical chaos, continuous POV, no artifacts.

Style: 8K IMAX, traditional hand-drawn 2D animation, animated on twos at 12 frames per second — each drawing held for two frames then replaced, choppy stepped motion cadence, visible pose-to-pose timing, distinct keyframe drawings with no smooth in-between interpolation, hand-painted oil-brush texture on every drawing, rough angular brushstrokes, brushstrokes shifting and redrawn from frame to frame, line jitter and boil between frames, no 3D render, no game engine, no CGI smoothness. DIRECTOR'S NOTES — non-negotiable for every frame and every cut: 1. WOLF REFERENCE IS ABSOLUTE — every wolf matches the design from <<<image_1>>> exactly: a lean steppe wolf, coal-black fur with a cold steel-blue sheen painted in rough angular brushstrokes, pale yellow eyes, low-slung predatory stance, head carried low. Rough painterly authorial style, angular and asymmetric, visible brushstrokes, matte, NOT Disney, NOT Pixar, NOT cute, NOT smooth glossy 3D fur. 2. THE PACK IS RANDOM, NEVER SYNCHRONIZED — this is critical. The wolves must all be different: different body sizes (a large lead wolf, mid-sized adults, two smaller younger ones), different poses, different head heights, each at a different phase of its stride, some sniffing low, some with heads up, some mid-step, some pausing, some lagging. NO two wolves in the same pose, NO mirrored or copied animals, NO synchronized movement, NO marching in step. Their motion is chaotic, staggered, organic, each animal moving on its own timing like a real wild pack. Asymmetric scatter, never a neat row, never evenly spaced. 3. PACK HIERARCHY — a clear large LEAD WOLF (the alpha) at the front, nose to the trail, the others fanned and strung out behind and below him at varying distances, the younger smaller ones trailing last. The lead acts first, the others follow raggedly, not all at once. 4. PREDATOR BEHAVIOR, NOT ATTACK — the wolves are tracking, not charging: nose to the snow reading the scent, ears swiveling, heads lifting to test the wind, slow deliberate creeping steps, pausing to listen, tense and focused. Quiet, controlled, ominous. Never running, never lunging. 5. BURAN NEVER PAUSES — violent buran with rolling white-out waves continuously in every frame, wind never drops, snow driving sideways. CHARACTER TAGS: - THE WOLVES = lean steppe wolves from <<<image_1>>>, a pack of six or seven, one large lead wolf and the rest varied in size. - THE FOREST = snowy forest slope location from <<<image_2>>>. Cinematography Lubezki / Deakins, aggressively handheld in the operator's own hands out in the storm — constant restless motion every frame, never still: jerky sway, the frame buffeted sideways by gusts, micro-jitter and shake throughout, breathing sway, occasional whip to recompose, the horizon never perfectly level, never gimbal-smooth, never tripod never dolly never crane never aerial, low to the ground near the wolves' eye level. The 12 principles of animation throughout: anticipation before each head lift and step, follow-through and overlapping on fur and tails lagging the body, slow in slow out on head turns, arcs, secondary action, exaggeration, solid drawing. Lighting: night, no light source, no moon, no stars, no rim, no contre-jour, no key light. Flat dim diffuse ambient grey glow only, no direction. Flat painted shapes, no gradient shading. Cold desaturated blue-grey palette. Atmosphere — violent buran continuous throughout every shot and cut: hurricane gusts driving snow horizontally, visibility roughly 3 meters, beyond dissolves into a pale blue-grey white wall, rolling waves of denser blizzard sweeping the lens swallowing the wolves briefly then revealing them, snow piling on their dark fur and backs, breath torn sideways from their muzzles, the fur on their backs and tails whipped by the wind, fine spindrift snaking across the snow between their legs. SHOT 1A — TIGHT MEDIUM CLOSE handheld on THE LEAD WOLF, low at his eye level among the dark trunks of THE FOREST. COMPOSITION: the lead wolf's head and shoulders fill the LEFT third, his muzzle down to the snow reading the scent, his dark angular mass against the pale white-out slope; empty foggy white forest fills the RIGHT, dark trunks as vertical masses framing the edges. He sniffs along the trail in the snow, nostrils working, breath puffing and torn sideways, then ANTICIPATION as his head slowly LIFTS on a slow arc, ears swiveling and pricking forward, pale yellow eyes opening toward the upper slope, he goes still and listens, muscles tense, the wind dragging his fur sideways. A rolling wave of buran sweeps across the lens, he is half-swallowed, re-emerges still listening. Handheld, jittering, shaken by gusts, low to the ground. Animated on twos. HARD CUT to SHOT 1B — EXTREME CLOSE handheld on THE LEAD WOLF's face from a low three-quarter angle, the angular rough-painted muzzle, the pale yellow eye the darkest-edged focal point, ears upright and twitching to catch sound through the wind. COMPOSITION: the face fills the frame off-center to the left, snow crusting on the fur and lashes, breath steaming and instantly torn away, white-out behind. He listens, a single ear flicks back then forward, the eye narrows, a low tension in the brow — he has caught something up the slope. Nostrils flare reading the air. Camera very close, breathing sway, shaken by a gust, snow whipping across the lens. Rough painterly brushstrokes, matte, never smooth. Animated on twos. HARD CUT to SHOT 1C — WIDE handheld among the trunks of THE FOREST, the whole pack revealed following the lead. COMPOSITION: the LEAD WOLF lower-LEFT in the mid-ground, head still up; behind and below him, strung out across the slope at random depths and distances, the rest of THE PACK emerges from the foggy white-out between the dark vertical trunks — six or seven wolves total, each DIFFERENT: one large, several mid-sized, two smaller young ones lagging; one with its nose down in the snow, one with head raised testing the wind, one mid-step with a paw lifted, one pausing half-hidden behind a trunk, one shaking snow off its back, one further back only a dark silhouette in the mist. They are scattered asymmetrically, never in a row, never evenly spaced, each at a different phase of movement, chaotic and staggered, no two alike, no synchronized motion. They creep forward and up the slope after the lead, slow and deliberate, dark angular shapes against the pale blue-grey storm, dark trunks slicing the frame. A rolling wave of buran sweeps across — several wolves vanish into white-out, then ghost back into view shifted, the pack reading as a loose living scatter of dark forms among the trees. Snow piling on their backs, fur and tails whipped sideways. Camera low, handheld, restless, buffeted by gusts. Animated on twos. Audio across all three shots: roaring buran wind continuously as dominant bed, fine hiss of driving snow, the wet snuffling of the lead wolf reading the scent, low breath and panting torn by the wind, the soft crunch of paws breaking snow crust at irregular intervals, ears and fur rustling, one low distant answering whine from a trailing wolf. No music. No dialogue. No subtitles. Constraints: three distinct shots with hard cuts between them not one continuous take — TIGHT MEDIUM CLOSE on the lead sniffing and lifting his head, EXTREME CLOSE on his listening face, WIDE revealing the whole scattered pack following — every wolf matches <<<image_1>>> exactly in the rough angular painterly style not smooth not Disney not cute not 3D, THE PACK IS RANDOM with different sizes poses head heights and stride phases, no two wolves alike, no mirrored or copied animals, no synchronized or marching movement, chaotic staggered organic motion, scattered asymmetrically never in a neat evenly-spaced row, a clear large lead wolf at the front with the others strung out behind and below at varying distances and the young ones trailing, the wolves are tracking and listening NOT attacking or running, frames composed off-center never centered with the pale white-out forest as active negative space, camera aggressively handheld low to the ground with constant shake and gust-buffeting never gimbal-smooth, buran rolling waves in every frame, no light source, flat painted shadow only, cold blue-grey palette, no 3D look.

Use @[character ref] as the only character and visual style reference. Start with action and end with action. A female gunfighter skates through a dark frozen arena at impossible speed, performing elegant ice-ballerina movements while shooting large numbers of enemies with a pistol. She attacks during pirouettes, spiral glides, sweeping turns, spins and rapid skating transitions without ever slowing down. Many enemies surround and rush her from all directions. Every skate movement leaves electric-blue ice trails, frost mist and icy spray. Every gunshot creates vivid crimson muzzle flashes and sharp graphic impact bursts. Keep the environment dark, moody and high-contrast, with bold red, blue and black posterized anime visuals. No sword, no slow motion, no static posing, no static enemies, full action throughout.

Use the attached THE STORM BREWER storyboard image as the exact reference. Create a 12-second 16:9 animated weather-wizard sequence that follows the 8-shot storyboard exactly. Preserve the same Pixar-style young weather-wizard — slate-blue long coat, leather apron, brass goggles, lightning-streaked silver hair — the same cliffside observatory workshop with copper tools and a wall of glowing weather-jars, and the same moody-blue / warm-amber color aesthetic throughout. Rules: - Follow the sequence exactly from 1 to 8 - One shot per panel, approximately 1.5 seconds each - No skipped steps, no extra steps beyond the storyboard - Maintain character and workshop continuity throughout - Emphasize the cloud catch, the lightning summon, the swirling bottling reveal, and the final lightning jar uncork Shot sequence: 1. Wizard arrives in the cliffside observatory and pulls his brass goggles down over his eyes — wide establishing shot of copper tools, oil lanterns, and the wall of empty bottles, with a stormy sky rolling past tall arched windows 2. Wizard leans out the window with a long copper net and scoops a swirling raincloud directly out of the sky — dramatic close-up on the net, captured cloud spiraling inside, raindrops trailing off the rim 3. Wizard tips the net over a massive copper bowl and pours the swirling cloud inside — side angle, vapor cascading down, tiny raindrops bouncing out and catching the lantern light 4. Wizard stirs the bowl with a long copper rod, the storm intensifies — full body shot, sleeves to the elbows, the first small sparks flickering up out of the bowl 5. Wizard pinches a tiny crackling lightning bolt between his thumb and forefinger and drops it into the bowl — extreme close-up on the hands, electric blue flash reflected in the goggles 6. Overhead shot of a row of glass jars being filled from a copper funnel — one fills with swirling clouds, one with falling rain, one crackles with caged lightning, each tagged with a paper label 7. Wizard slides the freshly corked jars onto a wall of hundreds of glowing weather-bottles — wide reveal of the magical inventory in pinks, blues, whites, and golds 8. Wizard holds up the lightning jar, pops the cork with his thumb, and a tiny crackling bolt arcs out the top — extreme close-up hero shot, his face lit pure electric blue, thunderclap, pure satisfaction Camera: - Wide establishing shot for the opener - Close-up on the cloud catch and the lightning summon - Side angle for the pour - Full body for the stir - Overhead for the bottling - Wide reveal shot for the shelving - Extreme close-up hero shot for the final lightning uncork Style: - Cool grey-blue storm light through tall arched windows, warm amber oil-lantern glow inside - Electric purple-white flashes during the lightning beats, mist and rain wisps swirling through the air - Pixar CGI vivid expressive animation, brass and copper highlights, glass with refracted light - Shallow depth of field on the close-up shots - Smooth satisfying cuts, electric and enchanting energy throughout Goal: A magical 12-second weather-wizard journey from storm-cloud catch to lightning crackle — electric, atmospheric, irresistible, and impossible to scroll past.

Style: Vertical 9:16 phone-shot UGC. Photorealistic — no 3D render, no game engine, no CGI skin. Camera: Modern smartphone front-facing camera. Slightly wide field of view, mild barrel distortion at the frame edges, imperfect digital depth separation. Micro handheld drift and small reframing corrections throughout — never a locked tripod, never a cine dolly move. Lighting: Practical sources only — warm amber floor lamp, warm string lights, soft window spill. Key from one side, deep blue-grey fill in the shadows, visible falloff into dark corners. Faint phone-sensor noise in the shadow areas. Color: 60:30:10 — warm amber room / deep blue shadow and costume / crimson and gold accent. Skin: Pore-level realism — vellus hair, faint capillary flush across the cheeks, natural facial asymmetry, real specular highlights on the nose and lower lip. No airbrushing, no waxy or plastic sheen. Acting: Natural creator energy — direct eye contact into the lens, conversational overlapping gestures, micro-pauses mid-sentence, small self-interrupting laughs, visible chest rise from breathing. She is talking TO the viewer, never performing at a camera. Physics: Gravity and inertia respected. Fabric drapes and settles with real weight; cape and hoodie lag a beat behind the body. Thin plastic film crinkles and holds creases. Correct contact shadows. No floating props. Composition: Subject centered to slightly low in frame, head-room typical of a hand-held phone. Continuity: Face, hair, wardrobe, room layout and prop placement identical across every shot. No identity drift. Technical: 30fps, 4K, natural motion blur. No jitter. No warping or extra fingers. Audio: Direct-to-camera dialogue, close and slightly roomy — a real bedroom, not a booth. Soft room tone underneath. No music. No subtitles, no on-screen text, no graphics, no UI overlays. Characters: MEI — "The Crimson Star" — @image1. Late-20s East Asian woman, long black hair with vivid crimson ombré through the lower half, loose natural waves, warm brown eyes, soft matte makeup. Wardrobe: cobalt-blue leather corset minidress with crimson piping and a crimson V-belt at the waist; cobalt-blue shoulder armor pauldrons trimmed in crimson and gold; cobalt-blue fingerless gauntlet gloves to the elbow; deep crimson knee-high heeled leather boots; full-length deep crimson cape with a thin gold edge, pooled on the rug behind her. CHEST EMBLEM — exactly one five-pointed crimson star, five equal points, solid crimson fill, thin gold outline, centered on the chest panel of the corset. Exactly five points, visible whenever her chest is not covered by the hoodie she is handling. Props: THE PACKAGE — translucent poly mailer, blue zip strip along the top now torn open, white shipping label on the front, plastic creased and crackling. It ends up discarded on the rug beside her knee. THE HOODIE — @image2. Oversized black cotton-fleece pullover hoodie densely scattered with small round silver rhinestones. Stone density is a gradient: heaviest across the hood and shoulders, thinning down the chest and sleeves, plain matte black at the ribbed hem and cuffs. Drop shoulders, boxy oversized fit, kangaroo front pocket, thin silver cursive script embroidered across the front edge of the hood. Heavy fleece — it hangs and swings with real weight. HOOD SCRIPT — per @image5, and this detail must be legible: a single line of flowing cursive script reading "AelfricEden", formed entirely out of small silver rhinestones set close together in continuous unbroken strokes, arcing across the front crown of the hood. The letterforms are joined handwriting-style — a large looping capital A, then lowercase "elfric" in a connected forward slant, then a large capital E whose crossbar sweeps long to the right, then lowercase "den" trailing off on a tapered final stroke. The stones forming the letters are slightly larger and packed far denser than the loose scattered stones around them, so the script reads as one bright continuous line against the open scatter. Keep the script sharp, unbroken, correctly spelled and readable. No garbled or invented letterforms, no doubled text, no mirrored text, no extra words, and no other text anywhere on the garment — the inner neck label is not visible in this shot. Scene: Set per @image4 — same apartment living room, same night, same lamp positions. CAMERA POSITION — the phone is still PROPPED on the low coffee table in exactly the position it held in the previous shot: lens at chest height, tilted slightly up, off-level by about two degrees, static. The camera does not move at any point in this clip — no handheld drift, no reframing, no push, no pan. Mei never holds or touches the camera. HANDS — both of Mei's hands are free and working the package and the garment throughout. She is not holding a phone or any device. No phone, no camera, no tripod, no ring light and no selfie stick appears anywhere in frame or in the background. Mei still sits cross-legged on the rug, dead center, cape pooled behind her, facing the propped camera. Framing is waist-up to knees-up, matching the previous shot exactly. This shot picks up the instant the previous one ended — both thumbs are already through the torn zip strip. No settling, no re-establishing. CUT 1 — Static shot from the propped phone, waist-up, no camera movement; she pulls the hoodie free: She tears the poly bag the rest of the way open with both hands and a hard plastic rip, reaches in and drags the folded hoodie out. It comes out heavier than her arms expected — her wrists dip, her shoulders compensate. She drops the empty bag onto the rug beside her knee where it settles and holds its creases. She says: "It's heavier than I thought it'd be." Her eyes are on the garment, not the lens, for the whole line. CUT 2 — Same static framing, she lifts the hoodie into the lamp light and rakes it: She grabs the shoulders with both hands and lifts the hoodie up into frame so it covers most of the shot, hood hanging toward camera and squared flat to the lens with the cursive script across the crown facing the camera dead-on and fully readable. She tilts it slowly through the amber key light from the floor lamp — no more than fifteen degrees one way, then back, keeping the hood panel flat to the lens the whole time so the script never skews, never curls away and never blurs. The dense rhinestone field across the shoulders and hood ignites in a rolling wave, hundreds of small warm specular points firing and going dark in sequence as the angle changes; the plain matte black at the hem stays dark and reads as fleece. Between the stones the fleece pile is fully readable. She says: "Oh my gosh — look at that. Look how much that catches the light." Her face is half-hidden behind the raised garment; one eye and the crimson ombré in her hair stay visible camera-left. CUT 3 — Same static framing, she gathers the hoodie up and rises straight toward the lens: The garment drops and she hugs it to her chest with both arms, the weight of it settling a beat after her hands stop. She is grinning, slightly wound up, breathing faster. Direct eye contact into the propped lens. She says: "Okay. I'm putting this on right now." On "now" she pushes up off the rug and strides directly toward the camera — the cape sweeping the rug behind her — until her torso fills and darkens the frame as she passes the lens. The camera stays exactly where it is; she moves, it does not. The shot ends on her body crossing the lens. Do not show her actually putting the hoodie on. End on the pass.

Style: Vertical 9:16 phone-shot UGC. Photorealistic — no 3D render, no game engine, no CGI skin. Camera: Modern smartphone front-facing camera. Slightly wide field of view, mild barrel distortion at the frame edges, imperfect digital depth separation. Micro handheld drift and small reframing corrections throughout — never a locked tripod, never a cine dolly move. Lighting: Practical sources only — warm amber floor lamp, warm string lights, soft window spill. Key from one side, deep blue-grey fill in the shadows, visible falloff into dark corners. Faint phone-sensor noise in the shadow areas. Color: 60:30:10 — warm amber room / deep blue shadow and costume / crimson and gold accent. Skin: Pore-level realism — vellus hair, faint capillary flush across the cheeks, natural facial asymmetry, real specular highlights on the nose and lower lip. No airbrushing, no waxy or plastic sheen. Acting: Natural creator energy — direct eye contact into the lens, conversational overlapping gestures, micro-pauses mid-sentence, small self-interrupting laughs, visible chest rise from breathing. She is talking TO the viewer, never performing at a camera. Physics: Gravity and inertia respected. Fabric drapes and settles with real weight; cape and hoodie lag a beat behind the body. Thin plastic film crinkles and holds creases. Correct contact shadows. No floating props. Composition: Subject centered to slightly low in frame, head-room typical of a hand-held phone. Continuity: Face, hair, wardrobe, room layout and prop placement identical across every shot. No identity drift. Technical: 30fps, 4K, natural motion blur. No jitter. No warping or extra fingers. Audio: Direct-to-camera dialogue, close and slightly roomy — a real bedroom, not a booth. Soft room tone underneath. No music. No subtitles, no on-screen text, no graphics, no UI overlays. Characters: MEI — off-duty look — @image3, face and hair per @image1. Late-20s East Asian woman, long black hair with vivid crimson ombré through the lower half, loose natural waves pulled out over the hoodie collar. Warm brown eyes, soft matte makeup, cheeks flushed, slightly out of breath from changing fast. Wardrobe: the oversized black cotton-fleece rhinestone hoodie from the previous scene, now worn — per @image2, stone density heaviest across the hood and shoulders, thinning down the chest and sleeves, plain matte black at the ribbed hem and cuffs; drop shoulders, boxy oversized fit, kangaroo front pocket, thin silver cursive script across the front edge of the hood, hood down at the top of the shot. Straight-leg mid-blue denim jeans, slightly loose, single cuff at the ankle. Plain white crew socks, no shoes. NO cape. NO shoulder armor. NO gauntlet gloves. NO corset. NO chest emblem — the costume is fully changed out of and no star of any kind appears anywhere in frame. Scene: Set per @image4 — same apartment living room, same night, same lamp positions, same rug and couch. The torn poly bag still lies discarded on the rug camera-right — leave it in shot, it sells the continuity. The camera is now hand-held again — she picked the phone up off the coffee table off-screen while changing, so this shot returns to front-camera selfie framing with the usual micro handheld drift. Mei stands in front of the couch holding the phone at arm's length in her left hand; her right hand is free and working the garment. This is a smash cut. She is fully dressed in the hoodie from the very first frame — no dressing action, no arms-through-sleeves, no transition. The clip opens with her already reacting. CUT 1 — Hand-held front-camera selfie, chest-up, she looks down at herself then up at the lens: First frame: she is already in the hoodie, hands spread flat against her own chest, looking down at the fit. She lifts her head into direct eye contact and her eyebrows go up. She says: "Oh my gosh. This is so nice." The amber lamp catches the stone field across her shoulders and throws warm specular points across the black fleece. The oversized fit reads clearly — drop shoulders sitting well past her actual shoulder line. CUT 2 — Same framing, she pulls the cuff and works the fabric: She pinches the inside of the opposite cuff between thumb and finger and rubs it, then pulls the hem out from her body and lets it fall back — the fleece drops with real weight and settles a beat late. Her eyes drop to her hands mid-sentence and come back up to the lens on the last word. She says: "It's so soft on the inside, that's what gets me." CUT 3 — She reaches up with both hands and pulls the hood over her head: She takes the hood in both hands and pulls it up over her head, adjusts it once at the temples, then sweeps the crimson-tipped hair out over the front of the hood so it falls down her chest. The rhinestone cursive script per @image5 sits across the crown of the hood above her forehead — foreshortened at this angle and not required to be fully readable, but it must read as one continuous bright line of script-shaped sparkle, denser and brighter than the scatter around it. Never as random dots, never as broken or garbled lettering, and never as any other word. The dense stone field across the hood ignites in the lamp light right beside her face — the brightest concentration of sparkle in the whole video sits inches from her cheek. She tilts her chin down, looks up into the lens from under the hood, small pleased smile. She says: "I need the hood up. Guys — look at this thing."

Style: Vertical 9:16 phone-shot UGC. Photorealistic — no 3D render, no game engine, no CGI skin. Camera: Modern smartphone front-facing camera. Slightly wide field of view, mild barrel distortion at the frame edges, imperfect digital depth separation. Micro handheld drift and small reframing corrections throughout — never a locked tripod, never a cine dolly move. Lighting: Practical sources only — warm amber floor lamp, warm string lights, soft window spill. Key from one side, deep blue-grey fill in the shadows, visible falloff into dark corners. Faint phone-sensor noise in the shadow areas. Color: 60:30:10 — warm amber room / deep blue shadow and costume / crimson and gold accent. Skin: Pore-level realism — vellus hair, faint capillary flush across the cheeks, natural facial asymmetry, real specular highlights on the nose and lower lip. No airbrushing, no waxy or plastic sheen. Acting: Natural creator energy — direct eye contact into the lens, conversational overlapping gestures, micro-pauses mid-sentence, small self-interrupting laughs, visible chest rise from breathing. She is talking TO the viewer, never performing at a camera. Physics: Gravity and inertia respected. Fabric drapes and settles with real weight; cape and hoodie lag a beat behind the body. Thin plastic film crinkles and holds creases. Correct contact shadows. No floating props. Composition: Subject centered to slightly low in frame, head-room typical of a hand-held phone. Continuity: Face, hair, wardrobe, room layout and prop placement identical across every shot. No identity drift. Technical: 30fps, 4K, natural motion blur. No jitter. No warping or extra fingers. Audio: Direct-to-camera dialogue, close and slightly roomy — a real bedroom, not a booth. Soft room tone underneath. No music. No subtitles, no on-screen text, no graphics, no UI overlays. Characters: MEI — off-duty look — @image3, face and hair per @image1. Long black hair with vivid crimson ombré through the lower half, loose waves, currently pulled forward over the front of the hood. Warm brown eyes, soft matte makeup, cheeks flushed. Wardrobe: the oversized black cotton-fleece rhinestone hoodie per @image2, hood UP, carried over from the previous shot — stone density heaviest across the hood and shoulders, thinning down the sleeves, plain matte black at the ribbed hem and cuffs; kangaroo front pocket, thin silver cursive script across the front edge of the hood. Straight-leg mid-blue denim jeans, cuffed once at the ankle. Plain white crew socks. NO cape. NO armor. NO gauntlets. NO chest emblem, no star of any kind anywhere in frame. Scene: Set per @image4 — same apartment living room, same night, same lamp positions. Mei sets the phone down on the low table at the top of the shot, giving a static, slightly off-level full-body wide of the cream rug and the grey-beige couch behind it. The torn poly bag is still on the rug camera-right. The amber floor lamp is camera-left, raking a hard key across the room at roughly waist height — this is the light that has to catch the back of the hoodie. CUT 1 — Frame swings and resettles static, full body; she plants center-frame: She steps to the low table, sets the phone down, and the frame lifts, wobbles and settles into a static slightly off-level full-body shot. She backs up onto the rug and plants center-frame, hood up, hands loose. Half-beat of stillness. She says: "And the rhinestones go all the way down the back too." CUT 2 — Same static framing, the turn: She turns a full 180 degrees to put her back to the lens, unhurried, and sweeps her hair over her right shoulder with her right hand to clear the back of the hood. She glances back over that shoulder into the camera. The raking amber key from camera-left crosses the back of the hoodie as she settles — the dense rhinestone scatter across the hood and shoulders fires in a rolling sequence with the shift of her weight, thinning progressively down the back and going fully matte black at the ribbed hem. The cursive script per @image5 wraps around the side of the hood crown and stays partly visible at this three-quarter-rear angle as a bright continuous curve of script-shaped sparkle. It is not required to be readable here, but it must not garble into random dots or resolve into different letters. Hold on the back. This is the money shot of the video; let it sit for a full beat with only her breathing moving the fabric. She says, still looking back over her shoulder: "Like — the whole back. That's the part that got me." CUT 3 — She turns front, hands into the pocket: She spins forward on the ball of one socked foot, the oversized hoodie swinging a beat behind her, and drops both hands into the kangaroo pocket. She rolls her shoulders inside the fit, chin tucked, and delivers the next line flat and factual — reporting, not selling, with a small shrug on the last word. She says: "I paid full price for this, by the way."

Style: Vertical 9:16 phone-shot UGC. Photorealistic — no 3D render, no game engine, no CGI skin. Camera: Modern smartphone front-facing camera. Slightly wide field of view, mild barrel distortion at the frame edges, imperfect digital depth separation. Micro handheld drift and small reframing corrections throughout — never a locked tripod, never a cine dolly move. Lighting: Practical sources only — warm amber floor lamp, warm string lights, soft window spill. Key from one side, deep blue-grey fill in the shadows, visible falloff into dark corners. Faint phone-sensor noise in the shadow areas. Color: 60:30:10 — warm amber room / deep blue shadow and costume / crimson and gold accent. Skin: Pore-level realism — vellus hair, faint capillary flush across the cheeks, natural facial asymmetry, real specular highlights on the nose and lower lip. No airbrushing, no waxy or plastic sheen. Acting: Natural creator energy — direct eye contact into the lens, conversational overlapping gestures, micro-pauses mid-sentence, small self-interrupting laughs, visible chest rise from breathing. She is talking TO the viewer, never performing at a camera. Physics: Gravity and inertia respected. Fabric drapes and settles with real weight; cape and hoodie lag a beat behind the body. Thin plastic film crinkles and holds creases. Correct contact shadows. No floating props. Composition: Subject centered to slightly low in frame, head-room typical of a hand-held phone. Continuity: Face, hair, wardrobe, room layout and prop placement identical across every shot. No identity drift. Technical: 30fps, 4K, natural motion blur. No jitter. No warping or extra fingers. Audio: Direct-to-camera dialogue, close and slightly roomy — a real bedroom, not a booth. Soft room tone underneath. No music. No subtitles, no on-screen text, no graphics, no UI overlays. Characters: MEI — off-duty look — @image3, face and hair per @image1. Long black hair with vivid crimson ombré through the lower half, loose waves falling forward over the front of the hood. Warm brown eyes, soft matte makeup, cheeks flushed, hood still up carried over from the previous shot. Wardrobe: the oversized black cotton-fleece rhinestone hoodie per @image2 — stone density heaviest across the hood and shoulders, thinning down the sleeves, plain matte black at the ribbed cuffs; kangaroo front pocket. Straight-leg mid-blue denim jeans, cuffed once at the ankle. Plain white crew socks. NO cape. NO armor. NO gauntlets. NO chest emblem, no star of any kind anywhere in frame. Scene: Set per @image4 — same apartment living room, same night, same lamp positions. Mei has picked the phone back up and sat down on the front edge of the grey-beige couch — hand-held front-camera framing, close, her face in the upper third. The amber floor lamp is behind her camera-left shoulder; the monstera reads soft and out of focus camera-right; warm string lights bokeh in the background. The torn poly bag is still visible on the rug in the lower background. CUT 1 — Close hand-held selfie, chest-up, she settles onto the couch edge: The frame swings up and resettles close on her face as she sits. Hood up, hair forward. She holds direct eye contact, warm and unhurried, and lets the energy drop a notch — this is the honest part of the video, not the excited part. She says: "It's on sale right now, though." Small nod on "though." CUT 2 — Same framing, the point: She lifts her right index finger into the lower-left area of frame and taps downward twice toward the bottom-left corner, past the edge of frame — pointing at something below the shot, not at the lens. Her eyes stay on the lens while her hand points. She says: "Click the orange cart right here to get it." IMPORTANT: do not render any cart, icon, button, badge, graphic, arrow, watermark or on-screen text anywhere in frame — the lower-left corner she points at is empty room. She is pointing at an interface element that is not part of the image. CUT 3 — The close: Her hand comes down out of frame. She leans in slightly toward the lens, eyebrows up, and gives a small closing smile with a single shoulder shrug. She says: "This is a must-have in your closet. Go." Then hold — a full beat and a half on her face, still smiling, blinking once, chest rising with a breath, hood framing her face and the stone field sparkling beside her cheek. The cursive script per @image5 sits across the crown above her forehead, heavily foreshortened — a bright unbroken line of script-shaped sparkle, not readable at this angle and not required to be, but never garbled into random dots or into different letters. The shot ends on the hold. No gesture flourish, no wave, no cut away.