AI fight generator. Starring you.

VERTICAL 9:16 15 seconds. ONE continuous handheld smartphone recording. Spectator at standing eye height beside a running lane. No cuts. No slow-mo. No captions. No music. No logos. CHARACTER LOCK — the person is exactly @image1. Preserve facial identity, hair, skin, age, build, gender presentation. Do not beautify or restyle. Use they / their / them only. @image1 = FACE only. Photocopy this person. Do not copy studio wardrobe from the lock photo. OUTFIT: casual fitted athletic clothes that suit the person in @image1, with white low-top sneakers. Their hair, exactly as in @image1, moves with the run. They are silent. Crowd noise only. SET: public plaza, dark-grey stone slabs, glass towers, trees, bright day. Dense spectators behind silver barriers. ONE enormous matte charcoal-grey curved ramp, several storeys high, starts off-screen right then enters during the pan. CAMERA: one phone lens, slight shake, tilted horizon, imperfect track, motion blur on the fast pan. Stay on the ground. Zoom ONLY after they land. 0.0–4.0s Already sprinting from left background toward camera. Crowd behind them. ~2–3s they rush past 1–2 metres away, filling the frame. Spectator pans right: front ¾ → side → rear ¾. They run toward the ramp. ~4s their foot hits the ramp’s lower curve. 4.0–6.9s Continuous run UP the steepening curve. Camera tilts up. They get small against the towering ramp. Last running footstep just below the lip launches them. Palms never touch the ramp. 6.9–9.9s Launch, several rapid end-over-end somersaults against sky and buildings, then they open out. ~9.9–10s BOTH sneakers hit the paving inside the barriers. Off-screen (~6.7–9.3s): “Oh my God, that was insane!” 10.0–12.4s Deep knee bend, recover, small step, upright by ~10.7, satisfied smile. Crowd erupts. Off-screen (~10.3–12s): “Whoa! That was crazy!” 12.4–15.0s Fast 3–4x smartphone zoom ~0.5s to a medium of them. ~13.2 they smile bigger, both arms up. Crowd still cheering. Off-screen (~12.7–13.6s): “Oh, no way!” End on raised arms. Same person every second. Crowd is behind barriers — extras, not a second hero. Live phone audio. Never a studio still as frame 0.

MMA — THE UNDERDOG. Ultra-photorealistic cinematic MMA fight in one octagon cage, about 30 seconds, fast rhythmic editing, alternating shot sizes and angles. @image1 is THE UNDERDOG; @image2 is THE FAVORITE. Keep each fighter's appearance exactly as their uploaded avatar; never swap the two identities and never change either avatar's appearance. Preserve their gender presentation exactly as shown. Use they / their / them only. Both wear hand wraps, a fitted rashguard and fight shorts, fully covered and decent, bare feet. THE UNDERDOG fights battered but unbroken; THE FAVORITE is cold and confident. THE FAVORITE dominates early with clean jabs, a cross, a low kick and a slow-motion looping hook that snaps THE UNDERDOG's head aside (a light spray of sweat and blood), and THE UNDERDOG staggers into the cage; THE UNDERDOG breaks a clinch and a cold focus settles in; THE FAVORITE commits to a heavy overhand right, THE UNDERDOG slips it and lands a SLOW-MOTION counter left hook and straight right, THE FAVORITE's confidence turning to shock, THE FAVORITE buckles and slides down the cage; a long cold close-up on THE UNDERDOG, then THE UNDERDOG stands over THE FAVORITE as the crowd erupts. TIMELINE (30s): 0–5s: THE FAVORITE dominates early with clean jabs and a cross. 5–8s: THE FAVORITE's low kick lands. 8–12s: the slow-motion looping hook snaps THE UNDERDOG's head aside (a light spray of sweat and blood) and THE UNDERDOG staggers into the cage. 12–16s: a clinch against the cage; THE UNDERDOG breaks it and a cold focus settles in. 16–19s: THE FAVORITE commits to the heavy overhand right and THE UNDERDOG slips it. 19–23s: the SLOW-MOTION counter left hook and straight right land, THE FAVORITE's confidence turning to shock. 23–25s: THE FAVORITE buckles and slides down the cage. 25–28s: the long cold close-up on THE UNDERDOG. 28–30s: THE UNDERDOG stands over THE FAVORITE as the crowd erupts, the shot holding to the final frame at 30s with no freeze and no early ending. Driving percussive score with constant crowd roar, dropping only on the two slow-motion beats, sharp clean strike thuds, no dialogue. Ultra-photorealistic, dramatic overhead arena lighting, film grain. Never swap identities, only a light realistic blood spray (no heavy gore), decent covered attire, no nudity, no warped faces or extra limbs, no on-screen text or watermark.

ASSET REFERENCING @avatar is the Hunter, the only person in the film. Gender, face, hair, skin tone, age and body proportions come entirely from @avatar and not from this text. ATTACHED IMAGES ARE MANDATORY AND FINAL FOR IDENTITY ONLY: wherever @avatar and this text disagree about the face, hair, skin tone, age or body of the Hunter, follow @avatar. For everything else - the clothing, the motorcycle, the harpoon, the creatures, the location, the framing, the camera and the light - this text is final and @avatar carries no information at all: ignore its clothing, background, framing and lighting completely. The hood always sits back far enough that the face of @avatar stays readable in every close shot. CHARACTER LOCK: in every shot the Hunter is exactly the person in @avatar - the same facial structure, eyes, eyebrows, nose, mouth, exact skin tone, hairline, hairstyle, hair length and colour, distinctive marks, apparent age, build and body proportions. Desert dust settles on the skin and hair of @avatar but never changes their colour. Preserve their gender presentation exactly as shown; the heavy riding gear does not masculinize, feminize or otherwise reinterpret it. No beautifying, no de-aging, no identity drift. WARDROBE LOCK (comes from THIS TEXT, never from @avatar): the Hunter of @avatar wears a long sand-brown hooded duster coat of heavy weathered canvas, torn and frayed at the hem, over dark scuffed leather body armour with a wide leather belt and pouches, a faded olive scarf around the neck, fingerless leather riding gloves, worn leather trousers and tall buckled riding boots, and a pair of scratched brass-rimmed goggles with cyan lenses pushed up on the forehead. The garment is fixed, only its drape adapts to the body of @avatar. The same coat, armour and goggles appear in every second of the film. HARPOON LOCK: a heavy steel harpoon launcher is strapped along the right forearm of @avatar. It fires one barbed steel harpoon that trails a thick steel chain; when it fires, crackling cyan electric light runs along the whole chain. It is always the same single launcher on the right forearm. MOTORCYCLE LOCK: one long chopper of ordinary motorcycle size, about two and a half metres long, built out of bones. Its frame is one huge bleached animal spine running from the headstock to the rear axle; a real ribcage curves around the engine like a cage; the front fork is two long bleached leg bones; the handlebars are a pair of curved horns; the headlight sits inside a large horned animal skull mounted at the front. The bones are dry, cracked and yellowed, lashed and bolted together with blackened iron plates and leather straps. A dark oil-stained engine sits inside the ribcage and a smooth bone-white fuel tank rests on the spine; both wheel hubs glow toxic green. Bone is the material the eye reads first in every shot of the motorcycle, from every angle; the only metal is the dark engine, the iron plates and the wheel rims. It carries no badges, no emblems and no lettering anywhere. It is always the same single motorcycle. THE SAND WORMS (from this text only): colossal eyeless desert worms with rough sand-coloured hide covered in overlapping dark armoured plates, and a round open mouth ringed with rows of long curved teeth. The smaller worms are about three metres thick; the largest one is about ten metres thick and rises forty metres above the road. Sand and asphalt pour off their bodies whenever they move. ONE-SENTENCE SUMMARY On a cracked desert highway at sunset, the Hunter of @avatar races a bone-framed motorcycle through a toxic desert as giant sand worms burst up through the asphalt, jumps a burning tank trailer, harpoons the largest worm, is dragged across the dunes, slingshots onto its armoured head, plants a charge and escapes on the motorcycle as the worm explodes behind - a photorealistic live-action blockbuster action sequence, 15 seconds of fast hard cuts. DETAILED PLOT 0-2s: Low tracking side shot from a camera car at full speed: the Hunter of @avatar leans over the bone motorcycle down the cracked desert highway, its bleached spine, ribcage and horned skull clearly readable in profile, the duster coat whipping back, the green hubs glowing. From the first frame the asphalt behind the rear wheel bulges and splits open in a long crack racing after the bike. 2-3s: Low frontal wide shot down the highway: a sand worm bursts straight up through the road ahead, its toothed mouth wide open, slabs of asphalt and a fountain of sand thrown into the air. 3-5s: Handheld chase shot from behind: the Hunter of @avatar swerves hard left and right between tilting slabs of collapsing asphalt, the rear tyre sliding and spraying grit, while the worm rises behind in a wall of dust and two columns of toxic green storm cloud tower on the horizon. 5-6s: Wide side shot: an old unmarked rust-red fuel tank trailer, with no cab attached, lies on its side across the road, burning. It stays there burning on the highway for the rest of the film. The motorcycle hits a tilted slab of asphalt like a ramp and flies over the flames, the Hunter of @avatar standing on the pegs, fire lighting the underside of the bike. 6-7s: Tight frontal close-up of the face of @avatar riding, dust on the skin, the hood edge flapping, the cyan goggles on the forehead catching the firelight, eyes locked on something above frame. A short push-in. No words. 7-8s: Low side angle: the largest worm rears up over the dunes; the Hunter of @avatar raises the right forearm and fires the harpoon. The barbed harpoon streaks up trailing the chain, cyan electric light racing along it, and bites deep between two armoured plates of the worm. 8-10s: Wide tracking shot across the empty dune crests, where the only things in frame are the worm, the chain, the Hunter of @avatar and the bone motorcycle: the chain snaps tight and the worm drags them across the dunes at terrible speed, the bike bouncing and sliding sideways, huge sprays of sand, the chain crackling cyan. 10-11s: The Hunter of @avatar locks the launcher and is yanked up off the seat, swinging through the air on the chain in a high arc toward the head of the worm; below, the motorcycle keeps rolling on alone across the flat sand toward the highway. 11-12s: Medium shot on top of the worm: the Hunter of @avatar lands hard on the armoured plates of its head, boots skidding, and grabs the edge of a plate with the left hand as the whole creature sways; the face of @avatar turns toward the camera, readable, jaw clenched against the wind. 12-13s: Close-up of the gloved right hand of @avatar pushing a fist-sized steel charge with a blinking cyan light deep into the gap between two armoured plates. 13-14s: Side shot: the Hunter of @avatar runs down the sloping back of the worm and drops from its flank straight onto the seat of the passing motorcycle, grabs the bars and twists the throttle, sand spraying from the rear wheel. 14-15s: Low frontal wide shot down the highway: the Hunter of @avatar rides straight toward the camera, the horned skull headlight of the bone motorcycle facing the lens, as the worm behind erupts into a gigantic fireball, a shockwave of sand and burning debris rolling outward and backlighting the rider in orange fire while the face of @avatar stays lit and readable in the glow. ADDITIONAL NOTES Photorealistic live-action blockbuster, filmed on location in a real desert with real stunt riding and real pyrotechnics, anamorphic lenses and camera-car rigs; every frame reads as real photographed footage, with natural skin texture, real sand, real dust, real fire and real motion blur. Colour stays burnt-orange sunset, sand beige and asphalt grey, with toxic green storm clouds on the horizon, the green glow of the wheel hubs and the cyan light of the harpoon chain as the only cool colours. One continuous geography: a straight cracked highway through low dunes, the burning tank trailer on the road, the dunes to the side of the road where the largest worm rises. Heat haze shimmers over the asphalt, the camera shakes hard on every impact, and sand, grit and asphalt chunks fly past the lens. The bone motorcycle, the harpoon launcher and the coat stay the same in every shot. The only vehicles in the whole film are the bone motorcycle and the burning tank trailer on the highway. The face of @avatar stays the same person in every close-up. Every hand has five fingers and grips what it holds. No badges, emblems or letters on any vehicle. Sound: the deep roar and rattle of the motorcycle engine, tyres skidding on grit, asphalt cracking and slabs crashing down, the low rumbling roars of the worms, sand pouring like water, the whoosh and crackle of the burning tank trailer, the sharp mechanical bang of the harpoon firing, the chain rattling and humming with electricity, heavy breaths, boots scraping on armoured plates, a high electronic beeping from the charge, and one huge thunderous explosion at the end. No BGM; generate only environmental sounds and action sounds. No dialogue and no voice-over. No subtitles.

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 share their face. 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. ASSET MAP: @image1 — the character (the driver). Photocopy @image1 for face, hair, skin, identity, and the wardrobe on the still (exactly the clothing and accessories visible in @image1). Do not restyle hair. Do not invent different clothing or accessories. Ignore any beige pad around the still — that pad is not a set. If any other person appears (the tactical soldier), they are never named, aged or gendered in the text — described only by their role and what they do. Only @image1's face is locked as an identity. Extra people in this story are beats, not copies of @image1. HANDS: hands are always anatomically correct with five fingers each, naturally proportioned, in every shot they appear. Create a cinematic photorealistic action-thriller sequence set in a modern downtown city at night, surrounded by tall skyscrapers, told as one clear cause-and-effect chain — not disconnected spectacle shots. SHOT 1 (0-4s): Close-up on the character driving, exactly as in @image1, one hand on the wheel, the other resting protectively on a small sealed metal case on the passenger seat. They glance anxiously into the rearview mirror, already uneasy — establishing that they are carrying something valuable and already suspect they’re being followed. Realistic skin texture, dramatic night lighting, tense expression. SHOT 2 (4-7s): Cut to a high aerial drone shot flying between the skyscrapers, following the character’s car through the busy night street below, then pulling back to reveal a heavily armored black tactical vehicle several cars behind, closing the distance with clear intent. The camera moves rapidly forward with cinematic motion blur, establishing that this vehicle is specifically hunting the character’s car. SHOT 3 (7-11s): Cut to a POV shot from inside the character’s car, focused on the rearview mirror. In the reflection, the tactical vehicle pulls up alongside, a tactical soldier visible inside, gesturing for them to stop. The character grips the case tighter and accelerates instead, confirming they won’t give it up. Keep the mirror reflection sharply detailed while the city blurs past. SHOT 4 (11-15s): Fast-paced aerial and street-level shots of the chase through downtown — the character weaving through traffic to shake the tactical vehicle, which rams a parked delivery truck while trying to cut them off, spilling cargo and forcing its own path toward a roadside fuel tanker idling at a loading zone. Realistic smoke, dust, and cinematic atmosphere. SHOT 5 (15-20s): The tactical vehicle, still trying to close the gap, clips the fuel tanker at speed. A powerful explosion erupts behind the character’s car — a bright orange fireball, thick gray smoke, sparks and debris — visibly caused by that collision, not appearing at random. The character’s car speeds clear of the blast radius; a final shot through the rearview mirror shows the fireball and stalled tactical vehicle receding into the distance as the character exhales, adrenaline still high but visibly safe. STYLE: Ultra-realistic physics, realistic vehicle movement, cinematic camera transitions, natural night lighting, shallow depth of field, high detail, realistic motion blur, blockbuster Hollywood action-movie cinematography, photorealistic, HDR, anamorphic lens, dramatic color grading, intense suspenseful atmosphere. AUDIO (native, synced, no BGM, no voiceover, no narration): engine roar and tire screech, wind rushing past the drone shots, the tactical vehicle’s engine snarling alongside, the crash into the truck, the tanker’s explosion with a deep boom and scattering debris, and the character’s tense breathing settling into relief in the final shot. No on-screen text, no subtitles, no watermark, no logos. Vertical 9:16. Exactly 20 seconds.

VERTICAL 9:16 16 seconds. Seedance 2.5. Cinematic action-comedy commercial. Hard cuts crane ↔ boardroom. Wide, macro, POV crosshairs, SLOW-MO. Photoreal. No on-screen text. CHARACTER LOCK: the sniper is exactly @image1. Preserve face, hair, skin, age, build. Preserve their gender presentation exactly as shown. Use they / their / them only. @image1 = FACE only. Photocopy this person. Sniper look: dark tactical clothes, futuristic silver arm armor. On a yellow construction crane over a modern city. @image2 = THIS product. Photocopy it. NEVER a different product. They load @image2 into the sniper rifle. The first shot is a LONG SLOW-MO of @image2 flying. A bucket or box of the same product sits beside them. SET A: yellow crane, city skyline. SET B: sleek glass boardroom. Several people in suits as extras, not extra heroes. One person in a blue top at the table receives the FIRST shot. 0.00–4.00 SETUP. Wide: they lie prone on the crane jib, load @image2 into the chamber. CU eye in the scope. POV crosshairs: distant boardroom. 4.00–10.00 FIRST SHOT — SLOW-MO HERO. Barrel fires. @image2 flies toward camera in LONG SLOW-MOTION. Smoke on the crane. Glass shatters in slow-mo. @image2 flies to the person in the blue top, who then catches it / holds it up, commercial, clean. 10.00–16.00 RAPID FIRE. Scope POV. They grab more of @image2 from the bucket, fire. More product through glass to the suited extras. End wide: table silent, product in hand. Action-comedy score. Same person as @image1. Same product as @image2. Never a studio still as frame 0.

Editing task: replace the Warrior in @video1 with the person in @image1. @video1 is the finished film. @image1 is the new Warrior: face, hair, hair colour, hairstyle, skin tone, age, gender presentation and body proportions come entirely from @image1. Preserve their gender presentation exactly as shown. Use they / their / them only. The clothing, background, framing and lighting of @image1 are ignored completely. Replace only the Warrior - the one unmasked fighter with the ember-orange cloak and the glowing ice axe - in every shot where the Warrior appears, including the close-ups of the face, the wide shots and the shots from behind - the lone charge across the frozen lake, the three axe strikes, the moment pinned by the crowd, the spinning break-out and the final blow that splits the lake - from the first frame to the last of the 30 seconds. The person from @image1 wears exactly the same frosted gunmetal armour, grey wolf-fur pauldrons, ember-orange cloak and ice axe as the Warrior in @video1; the armour is fixed, only its fit adapts to the body of @image1. Frost settles on them but never changes their hair, hair colour or skin tone from @image1. Keep everything else in @video1 exactly as it is: every shot, every cut and its timing, every camera move, every action and pose of the Warrior, the frost soldiers, the Frost Monarch, the frozen lake, the ice spire, the blizzard, the light and colour, and the soundtrack. Follow the reference clip's timing from 0s to the final frame at 30s, with no freeze and no early ending. No subtitles.

ASSET REFERENCING @image_2 is the Mentor (the second attached image), the only person ever seen in full in this film. Gender, face, hair, skin tone, age, voice and body proportions come entirely from @image_2 and not from this text. @image_1 is the location, the first attached image: the black obsidian bridge, the lava sea and the crimson fortress-palace on the volcano. ATTACHED IMAGES ARE MANDATORY AND FINAL FOR IDENTITY ONLY: wherever @image_2 and this text disagree about the face, hair, skin tone, age or body of the Mentor, follow @image_2. For everything else - the robes, the swords, the framing, the camera and the light - this text is final and @image_2 carries no information at all: ignore its clothing, background, framing and lighting completely. The location comes from @image_1 and from this text. CHARACTER LOCK: in every shot the Mentor is exactly the person in @image_2 - the same facial structure, eyes, eyebrows, nose, mouth, exact skin tone, hairline, hairstyle, hair length and colour, distinctive marks, apparent age, build and body proportions. Preserve the gender presentation of @image_2 exactly as shown; the robes and the armour do not masculinize, feminize or otherwise reinterpret it. Real skin texture, their own hair exactly as in @image_2 stirring in the hot wind, clear but restrained micro-expressions. No beautifying, no de-aging, no identity drift. WARDROBE LOCK (comes from THIS TEXT, never from @image_2): the Mentor of @image_2 wears one flying-warrior outfit: a long deep-crimson layered silk robe with narrow sleeves, a black lacquered leather chest guard and a wide black waist sash trimmed with thin gold edging, black leather bracers and soft black boots. The garment is fixed, only its drape adapts to the body of @image_2. The same crimson robe in every second of the film; the Mentor never changes clothes. THE HERO (from this text only): the camera IS the eyes of the Hero, at natural standing eye height. The Hero is never seen from outside; only the Hero's own hands, forearms and boots enter the frame: dark crimson sleeves, black leather bracers, black boots. SWORD LOCK: exactly two swords exist. The Hero's sword: a straight double-edged sword about one metre long in a black lacquered scabbard with plain gold fittings. The Mentor's sword: the same shape in a dark red scabbard. When a sword flies, its dark steel blade glows molten orange along both edges and leaves a thin trail of fire; it stays the same length and flies flat, point forward, and a standing person rides it with both feet on the blade. Every hand has five fingers. ONE-SENTENCE SUMMARY On the obsidian bridge above the lava sea in @image_1, the Mentor of @image_2 throws the Hero's sword into the lava to force the Hero's first sword flight, then leads a fast first-person chase under the bridge, through the lava falls and up the crimson fortress-palace - an ultra-realistic cinematic fantasy film, 15 seconds, one continuous first-person POV, English dialogue. DETAILED PLOT 0-1s: First-person POV at walking pace on the black obsidian bridge of @image_1, the lava sea glowing far below on the left, the crimson fortress-palace rising on the volcano to the right. The Mentor of @image_2, walking beside the camera in the crimson robe, stops abruptly, turns to face the lens and holds out one open hand. The Mentor of @image_2: "Give me your blade." 1-2s: The Hero's own hands enter the frame and hand over the sheathed sword in the black scabbard. The Mentor of @image_2 takes it and in one movement turns and throws it over the black-stone balustrade out into the void. 2-3s: The camera rushes to the balustrade and looks straight down; the black sword tumbles toward the lava sea far below. The Hero, shocked: "What are you doing?!" Far below, the falling sword bursts into flame, slows, turns flat and hangs in the air above the lava on a trail of fire. 3-4s: The camera lifts to the Mentor of @image_2, standing at the balustrade, who looks straight into the lens. The Mentor of @image_2: "Fire doesn't wait. Neither do you." The Mentor's own sword slides out of the dark red scabbard by itself, ignites along both edges, flips flat and hovers a few metres below the balustrade. 4-5s: The Mentor of @image_2 steps lightly up onto the balustrade, balance shifting realistically, glances back at the lens. The Mentor of @image_2: "Jump." Then steps backward off the bridge. 5-6s: The camera lunges to the balustrade and looks down: the Mentor of @image_2 drops in a short free fall, lands with both feet on the flaming sword, knees bending to absorb the impact, and shoots forward. At the same moment the Hero's own sword rises back out of the depth on its fire trail and hovers flat and glowing just below the balustrade in front of the camera, clearly visible. 6-7s: The Hero climbs over the balustrade and steps off. A short, real fall; the Hero's black boots slam onto the glowing blade of the Hero's own sword, visible in the frame under the boots, the view dips hard, sways left and right as the body corrects its balance, one small shake, then steadies. 7-8s: The sword accelerates after the Mentor of @image_2, who flies a few metres ahead in the crimson robe and shouts back over one shoulder: "Don't look at the lava. Look at me!" 8-9s: Following the Mentor, the camera banks hard left along the outside of the huge obsidian arch; the balustrade and the underside of the bridge rush past overhead with strong parallax, carved black stone sliding over the lens. 9-10s: The flight drops under the arch and bursts through a thin glowing curtain of molten lava pouring off the bridge edge, sparks spraying past the lens, then skims low over the lava sea between tall columns of fire erupting on the left and right. 10-11s: The Mentor of @image_2 pulls up first along the crimson fortress-palace; the camera follows in one smooth climbing arc: colossal red pillars, burning bronze braziers and long balconies sweep past close on the right, the deep lava sea always visible on the left, speed strong but the space clear. 11-12s: The Mentor of @image_2 banks around a huge upswept pagoda eave and disappears for a moment behind the palace corner; the camera banks after the Mentor along the same line. 12-13s: The camera bursts out of the shadow into the blood-red sunset light: the Mentor of @image_2 now flies parallel a few metres away. The Hero's view is finally steady. The Mentor of @image_2 looks across, sees the Hero is no longer wobbling, and one corner of the mouth lifts. The Mentor of @image_2: "See? You didn't burn." 13-14s: The Hero, breathing hard: "You call that teaching?" The Mentor of @image_2 laughs once, softly, and leans forward low over the blade. The Mentor of @image_2: "Lesson two - catch me." 14-15s: The Mentor's sword surges ahead on a long trail of fire between the fortress terraces and the lava sea, toward the glowing crater of the volcano; the camera leans forward and accelerates after the Mentor of @image_2. The film ends on this fast first-person flight toward the fortress lit by the red sunset. ADDITIONAL NOTES One continuous first-person POV from the Hero's real eyes for the whole 15 seconds, never a third-person view, never a floating or drone-like camera: every move - walking, rushing to the balustrade, looking down, falling, landing on the sword, losing and finding balance, banking, diving, climbing - has real start, weight, inertia, correction and settling. The story reads on first viewing: the Mentor takes the Hero's sword and throws it off the bridge, the Hero jumps for the first time, the Hero flies, the Mentor starts lesson two. The sword flight starts at 6 seconds and fills the rest of the film. The bridge, balustrade, lava sea, fortress-palace, eaves and terraces of @image_1 keep stable, continuous geography; the people are tiny against the architecture. The face of @image_2 stays the same person in every close look. Only the Hero's hands, forearms and boots ever appear from the Hero. Every word of dialogue is spoken aloud in English, lips in sync; the Mentor's voice fits @image_2, the Hero's voice is a young adult voice, breathless in the air. No banners, emblems, flags or letters anywhere. Look: ultra-realistic live-action fantasy cinema, 24 fps motion, real exposure, Kodak film grain, crimson, black, gold and molten orange palette, warm lava light from below, embers drifting in the air. Sound: hot wind roaring past the ears, the deep rumble and bubbling of the lava sea, the whoosh and crackle of the flaming swords, the thud of boots landing on steel, fire columns erupting, the hiss of the lava curtain, heavy breathing, and the spoken dialogue. No BGM; generate only environmental sounds, action sounds and the dialogue. No subtitles.

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 share their face. 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. ASSET MAP: @image1 — the character (the frightened protagonist). Photocopy @image1 for face, hair, skin, identity, and the wardrobe on the still (exactly the outfit and any jewellery @image1 shows). Do not restyle hair. Do not invent a different outfit or jewellery. Ignore any beige pad around the still — that pad is not a set. If any other person appears (the tactical soldier), they are never named, aged or gendered in the text — described only by their role and what they do. Only @image1's face is locked as an identity. Extra people in this story are beats, not copies of @image1. HANDS: hands are always anatomically correct with five fingers each, naturally proportioned, in every shot they appear. Create a cinematic photorealistic action-thriller sequence set in a modern downtown city surrounded by tall skyscrapers, at dusk with the city lights just coming on. SHOT 1 (0-3s): Extreme close-up of the character’s frightened face — the person exactly as in @image1 — wet eyes, realistic skin texture, dramatic dusk lighting, intense emotional expression. SHOT 2 (3-7s): Cut to a high aerial drone shot flying between dense skyscrapers, looking down at a busy city street with cars, pedestrians, traffic markings and realistic urban detail. The camera moves rapidly forward between the buildings with cinematic motion blur. SHOT 3 (7-12s): Cut to a dynamic POV shot from inside a moving dark sedan, focusing on the side-view mirror. In the mirror reflection, a heavily armored futuristic black tactical vehicle drives aggressively alongside the car, with a tactical soldier visible inside — described only by role and action, never named or identified. Keep the mirror reflection realistic and sharply detailed while the surrounding city moves with natural motion blur. SHOT 4 (12-16s): Fast-paced aerial and street-level shots showing the armored vehicle chasing the sedan through the downtown streets, including a brief stretch through an underpass tunnel lit by sodium lights. Realistic smoke, dust, damaged vehicles and cinematic atmosphere. SHOT 5 (16-20s): End with a dramatic top-down aerial shot looking directly at a city intersection as the sedan moves through traffic and a powerful explosion suddenly erupts nearby, producing a bright orange fireball, thick gray smoke, sparks and debris. STYLE: Ultra-realistic physics, realistic vehicle movement, cinematic camera transitions, natural lighting, shallow depth of field, high detail, realistic motion blur, blockbuster Hollywood action-movie cinematography, photorealistic, HDR, anamorphic lens, dramatic color grading, intense suspenseful atmosphere. AUDIO (native, synced, no BGM, no voiceover, no narration): engine roar and tire screech, wind rushing past the drone shots, the tactical vehicle’s engine snarling alongside, distant sirens, the explosion’s deep boom with debris scattering, and the character’s frightened breathing in the close-up. No on-screen text, no subtitles, no watermark, no logos. Vertical 9:16. Exactly 20 seconds.

CINEMATIC ACTION — NINJA SHURIKEN PARKOUR, hyper-realistic, full realtime speed. @image1 is the ninja hero, identity kept exactly from the reference the whole time: a sleek modern ninja in fitted dark tactical garb with a half-face wrap that leaves the eyes and upper face clearly visible; agile, fierce, fully covered and decent. @image1 runs, leaps and performs acrobatic parkour flips across rooftops and a moody neon back-alley at night, pursued by a group of solid, real, opaque human figures in ragged dark clothing — full flesh-and-blood bodies that catch the neon light and cast real shadows. @image1 fights back by hurling steel throwing stars (shuriken), drawing them fast and flinging them with sharp arm snaps; the spinning steel stars streak through the air trailing faint light and strike the pursuers, dropping them back, blending parkour with combat. BEATS (20s): 0–4s: @image1 sprints and vaults across a rooftop, the pursuers chasing. Camera: tracking. 4–8s: @image1 spins mid-run and flings two shuriken that spin through the air and hit two pursuers who stagger back. Camera: whip with the throw. 8–12s: a wall-run and a flip down into a neon alley, drawing more stars. Camera: follow the descent. 12–16s: a fast combat flurry — @image1 throws a spread of shuriken in different directions, each striking a closing pursuer, sparks flying. Camera: handheld combat. 16–20s: @image1 lands in a low crouch, one last shuriken spinning on a fingertip, then rises and turns to camera, fierce and composed. Camera: slow push-in. LOOK: hyper-realistic cinematic action, 35mm handheld, shallow depth of field, moody neon night in teal and magenta, realistic spinning steel shuriken with motion trails, natural motion blur, film grain, photoreal throughout, clean frame with no on-screen text or logos. (music only, mixed loud and full so the music dominates from the first frame: tense driving cinematic action percussion) <whooshing throwing stars, metal shings and impacts, fast footsteps and vaults>

VERTICAL 9:16 18 seconds. ONE continuous take. Camera LOCKED: full-body, tripod, no pan, no tilt, no cut. @image1 is THE PERFORMER — photocopy FACE and hair only, exactly as in @image1. Do not keep any clothing or accessories from the still. Wardrobe from the scene: black loose martial-arts uniform, white undershirt, white socks. Grey brick wall, paved outdoor ground. No other people. No on-screen text. They are already in the first pose at frame 0 — never a studio headshot. 0:00–0:02 Balance on RIGHT leg. LEFT leg high, foot to the ceiling. Right arm up, palm out. Left arm down, palm down. Slowly lower the left leg. 0:02–0:04 Left leg coming down. Both arms cross at the chest, then open: right arm forward, left arm back. Still on the right leg. Left knee bent, foot pointing down. 0:04–0:06 Swift high left kick straight up. Right arm above the head, left hand to the waist. Hold a beat, then left knee bends again. 0:06–0:10 Left knee bent, deep squat on the right leg. Arms cross at the chest, then right arm up, left arm forward, palms out. Hold the low squat. 0:10–0:13 Rise on the right leg. Arms cross again. Left leg extends forward, knee bent. Left arm up. 0:13–0:18 Left leg kicks straight out, toe pointed. Right hand to the face, palm out. Left hand at the waist. Slowly sink into a deep squat on the right leg with the left leg still straight out. Hold. Intense stare into the lens to the end. Precise, controlled, one breath. No freeze-as-JPEG except the designed holds.

30 seconds, 9:16. ONE continuous take. action-cam forehead POV then selfie. Photoreal. Adult. PERSON LOCK: @image1 is the hero — photocopy their FACE, hair and any facial hair exactly as shown. Preserve their gender presentation exactly as shown. Use they / their / them only. IGNORE whatever clothes are in @image1. Face only appears in the last 8 seconds. OUTFIT the whole clip: plain swim shorts and a plain tank, barefoot. Canyon: PNW pine forest, basalt hexagonal cliffs, turquoise spring pool, two waterfalls, submerged logs, pink-and-white tube. 0–2s POV looking down from the cliff. Bare feet on the ledge, swim shorts at the bottom of frame. Hands at sides. Toes curl. Right hand tosses the pink-and-white tube; it falls to a speck on the water. 2–4s Nervous tilt down, exhale, two steps back. Spoken, hyped: "Three, two, one!" 4–7s Slow-mo takeoff. Two running steps, last foot leaves the ledge. Arms out. Canyon opens below. 7–13s Real-time fall. Frame wobbles. "Woooooo!" Walls streak. Water rushes up. Pencil dive. 13–14s Feet-first splash. Bubbles swallow the lens. 14–17s Underwater, sun shafts, logs, one breaststroke, kick up. 17–19s Surface. Grabs the tube. 19–22s Bobbing, tilt up at the cliff. Spoken, called up to a friend who stays off-screen and never appears: "Yeah, friend! That was insane!" 22–30s Still in the water, they TURN the camera around to themselves, holding it like a selfie. The face of the person in @image1 now in frame, wet, excited, same energy. They look into the lens. Spoken, excited: "Have to try this again!" HOLD. AUDIO: wind, waterfall, splash, spoken English. Chill electronic without lyrics. No overlay type. NEGATIVE: clothing from @image1, overlay captions, extra people, 16:9.

15 seconds, 9:16. Photoreal. Follow the shot plan below exactly. The person in @image1 is PLANTED — they do not walk. ONE shot. Slow push-in from full-body to close-up. No hard cuts. PERSON LOCK: @image1 is the hero — photocopy their face and hair colour exactly as shown. Preserve their gender presentation exactly as shown. Use they / their / them only. Their own hair exactly as in @image1, neatly styled; dressed in smart, plain office clothes that suit them. Adult. KEEP this outfit. Same person every frame. No turtleneck. No pleated skirt. SETTING: modern glass-and-wood CORPORATE OFFICE corridor. Desks, glass rooms, warm daylight. NOT a school. No lockers. No ninjas. No weapons. EXTRAS: colleagues in BLACK SUITS, approaching with papers, folders, pens, coffee cups. 0.0–2.0s LOCKED full-body. The person in @image1 STANDS center, facing camera, arms at their sides, calm. Black-suited colleagues rush AT them at NORMAL SPEED with papers. 2.0–4.0s They raise their RIGHT HAND, OPEN PALM toward camera — a clear stop-gesture. Still planted. Still calm. 4.0–9.0s SLOW MOTION. Papers, pens, folders, coffee cups FLY around them. Colleagues STEP BACK, thrown off balance. Their hair lifts and drifts as if in wind. Camera SLOWLY PUSHES IN. Palm still toward camera. 9.0–11.5s Tighter. They switch to a raised INDEX FINGER. Hair still in slow-mo wind. Debris still floating. Extras receding. 11.5–15.0s CLOSE-UP of their face. Chaos falls out of frame. They lift their LEFT HAND to their left ear / hair, calm. HOLD. AUDIO: office, whoosh on the palm, paper flutter. No speech. NEGATIVE: walking, school, ninjas, batons, turtleneck, hair recast, overlay type, extra celebrity faces, tracking-back walk.