GPT Image 2 prompts

Cute stylized {argument name="subject" default="animal"} made from smooth river stones, stacked organic forms, matte mineral texture, {argument name="action" default="sitting quietly"}, calm zen atmosphere, rounded proportions, natural {argument name="color tones" default="earthy"} tones, minimal composition, soft directional lighting, collectible art object aesthetic.

High-fashion editorial studio photoshoot of a stylish young woman with {argument name="hair style" default="short wavy black hair"}, wearing {argument name="top" default="oversized dark olive cable-knit sweater"}, black cargo pants, chunky black leather combat boots, and {argument name="eyewear" default="orange-tinted sunglasses"}. She is posing in a deep squat/crouching position, holding a structured burnt-orange leather handbag. Minimalist monochromatic orange seamless backdrop and floor, warm color palette, luxury fashion campaign aesthetic, confident expression, soft cinematic studio lighting, sharp focus, premium fashion magazine photography, ultra-detailed textures, natural skin, modern streetwear meets luxury style, full-body composition, front view and side profile angles, shallow depth of field, high-end commercial photography, 85mm lens, photorealistic, 8K quality.

Transform the entire image into a single {argument name="style" default="Decorative Folk Flat Illustration"} with Doodle elements. Use a {argument name="color palette" default="bold and playful color palette"}, completely different from the original image. Simplify all details into clean, flat shapes with a handmade, slightly imperfect feel, as if drawn on a sheet of {argument name="canvas" default="white paper"}. The overall style should look cute, childlike, and whimsical

{argument name="platform" default="Netflix"} original series poster, close-up portrait, intense expression, cinematic color grading, dramatic atmosphere, subtle film grain, premium streaming-series aesthetic, shallow depth of field, ultra-detailed face, award-winning photography, 8K.

Editorial fashion portrait, woman wearing {argument name="clothing" default="flowing red bell-sleeve wrap dress"}, dramatic hard rim lighting from camera-left casting golden edge glow on fabric, direct intense eye contact with camera, parted lips, confident expression, wind-blown sleeve fabric captured mid-motion with natural fabric ripple, hair strands lifted by wind, deep saturated red shadows contrasted against {argument name="lighting highlight" default="warm blown-out gold highlights"}, asymmetric tight crop with one sleeve extending off-frame, cinematic color grading, high dynamic range, shot on {argument name="camera" default="Hasselblad medium format"}, 85mm lens, f/1.8 shallow depth of field, warm amber gradient backdrop with subtle atmospheric haze and soft lens flare bottom-left corner, skin texture realistic with subtle highlight sheen, professional studio fashion photography, ultra detailed

Create a crisp, clean infographic storyboard poster for {argument name="business name" default="THE CINNAMON ROLL BAKER"}. Wide 16:9 layout, white background, black borders, bold black typography, {argument name="style" default="premium Pixar 3D"} stylized rendering, bright warm colors — golden baked dough, rich cinnamon

A highly detailed 3D animated character design sheet of a cheerful little girl named Luna and her bunny companion Bun, presented in a clean professional animation studio turnaround layout, Pixar-style, DreamWorks-style, soft pastel aesthetic, white

Create a cinematic anime-style vertical tram stop scene at golden-hour sunset after rain. A glossy black modern streetcar decorated with elaborate metallic gold chrysanthemum and floral patterns is stopped at a traditional Japanese station named {argument name="station name" default="Yuhidai 1-chome"}. The tram destination sign should show Japanese text with small English beneath it, reading {argument name="tram destination" default="KINKO NO SAITO - YUHIDAI"}, and the side of the tram should read {argument name="tram line name" default="Kinko Tram Line"}. In the foreground, an elegant young anime woman with very long flowing {argument name="hair color" default="black"} hair steps down from the tram, wearing an ornate black-and-gold kimono with floral embroidery, white tabi socks, sandals, and a decorative gold hair ornament; she carries a small matching handbag and looks calm and slightly mysterious. Include exactly two visible human figures: the kimono-clad woman outside the tram and the uniformed tram driver seen through the front windshield. The tram has warm glowing interior lights, reflective windows, bright headlights, visible pantograph and overhead wires, and an additional painted anime-girl portrait on the tram front as part of the decoration. To the right, place a tall illuminated station sign with Japanese vertical text and small English text underneath reading "YUHIDAI 1-CHOME", beside a traditional wooden building with lit windows and tiled roof eaves. Use a dramatic orange sky with glowing clouds, low sun near the horizon, wet reflective pavement and rails, small lanterns, golden falling petals, lens flares, and rich amber highlights. Make the image highly detailed, polished, atmospheric, and luxurious, blending Kyoto-like traditional architecture with a sleek fantasy tram design. No extra people, no modern cars, no watermark, no logo clutter, avoid distorted text beyond the specified signage.

Create a crisp, clean infographic storyboard poster for {argument name="project title" default="THE MURALIST"}. Wide 16:9 layout, white background, black borders, bold black typography, premium Pixar 3D stylized rendering, explosive vivid colors — {argument name="primary color" default="electric cobalt blues"}, {argument name="secondary color" default="vibrant sunflower"}

Goal: Create a clean GitHub README-style landing page screenshot promoting {argument name="project title" default="GPT Image 2 Prompt Gallery + Agentic Skill + CLI"}, with a centered title area and a large fantasy-anime hero banner underneath.

Canvas: 890×650 px webpage screenshot on a white background, with generous margins and a thin light-gray horizontal divider below the title. Use a crisp browser/documentation aesthetic, no browser chrome.

Layout: Top-left has a small black chain-link anchor icon. Centered at the top is the large bold headline “GPT Image 2 Prompt Gallery + Agentic Skill + CLI”. Under it, place an italic two-line subtitle: “OpenAI GPT Image 2 prompt gallery, image prompt library, agentic skill, and CLI — curated, copy-paste prompts and runnable examples for skill-capable agents.” Below that, centered language links read “English · 中文” in blue underlined link style. Under the language links, place exactly 6 small rectangular status badges in one row: “License MIT”, “PRs welcome”, “model gpt-image-2”, and “python ≥3.11”, with the first two labels split into dark-gray label + green value, the model badge split dark-gray label + purple value, and the python badge split dark-gray label + blue value. Below, place exactly 3 centered ranking badges in one row: “oosmetrics Top 1 Agents”, “oosmetrics Top 1 LLMs”, and “oosmetrics Top 1 CLI”, using dark-gray label segments and muted blue/purple value segments.

Hero banner: Place one large wide rectangular illustration beneath the badges, spanning almost the full content width with small white margins. The banner is a richly detailed fantasy workshop scene in warm cinematic lighting, anime/chibi style, with a black brush-stroke sign across the upper center reading {argument name="banner title" default="GPTImage2Skill"} in large rough gold lettering. The background shows a magical cave workshop with glowing UI panels, floating icons, scrolls, prompt sheets, potion-like colorful orbs, tools, and a mountain landscape opening in the distance.

Characters and objects in the hero banner: Include exactly 8 chibi fantasy characters arranged across the banner, each holding or presenting image-generation artifacts: 1 boy on the left holding a parchment and raising a fist, 1 purple-clothed character at bottom-left holding a dramatic anime image, 1 green-clothed character near lower-left holding a diagram/spec sheet, 1 elderly robed character near center-left, 1 small red-clothed character in the center, 1 green/yellow character near center-right holding a phone-style UI sheet, 1 smiling orange/green character near lower-right holding a realistic bird photo, and 1 blue/green character at far right holding a poster and a dashboard panel. Add several additional partial humanoid figures behind them if needed, but the 8 foreground chibi characters must remain visually dominant. Render the humanoid faces as intentionally blurred or masked soft tan rectangles, as if anonymized, while keeping costumes, poses, hands, and held images sharp.

Visual style: Combine clean documentation page typography with a vibrant high-detail fantasy anime illustration. The page area should be flat and white; the hero image should be saturated, warm, magical, and highly detailed with glowing cyan and gold accents. Keep all visible text legible where specified, avoid adding extra paragraphs, watermarks, logos, or browser controls.

Create a realistic cinematic photograph of a {argument name="subject" default="Japanese teenage schoolgirl"} standing indoors in a quiet school hallway beside a large window, shown from the waist up in side profile as she gazes outside. She wears a dark navy sailor-style school uniform with a pleated skirt, white double-striped sailor collar, and a muted red necktie; her long straight black hair falls past her shoulders. Her face is intentionally obscured by a soft square mosaic blur centered over the facial area, preserving anonymity. The hallway behind her is dim and slightly out of focus, with a door and frosted rectangular window panel visible in the background. The right side of the image is dominated by tall white-framed windows and a pale windowsill, with a rainy or overcast outdoor view of distant hills, buildings, greenery, and a faint reflection of the girl in the glass. Use natural soft daylight, muted cool colors, shallow depth of field, subtle film grain, contemplative mood, photorealistic 35mm documentary style, vertical square composition, no text, no watermark.

Goal: Create a premium cinematic diptych poster for an ancient Chinese fragrance and beauty product campaign, showing exactly 2 vertical poster panels side by side on a white background with a narrow white gutter between them. Use a dark, luxurious, historical palace aesthetic with dramatic low-key lighting, gold typography, and editorial perfume-ad composition.

Canvas: Square overall image containing two tall portrait posters, each with a thin white border. Left poster is warm amber-gold and smoky; right poster is cool moonlit blue-black and misty. Both posters should look like high-end Chinese cosmetic advertisements with ornate calligraphy and small seal stamps.

Left panel: Create a Tang-style dancer fragrance poster titled {argument name="left main calligraphy" default="赵飞燕"}. The scene is a dim imperial interior with carved wooden screens, candlelight, incense smoke, and silhouettes of musicians in the background. A graceful young Chinese woman in ancient court costume sits or reclines in the center-right foreground, face obscured by a soft square blur, arms raised above her head in an elegant dance pose. She wears an elaborate gold floral hairpiece with dangling chains, jade and pearl ornaments, bare shoulders, a translucent green-gold silk wrap, and delicate jewelry. In the lower-left foreground place exactly 3 product objects: 1 ornate black-and-gold cloisonné perfume vase emitting smoke, 1 matching round decorated compact or lidded perfume case, and 1 open small cream bowl with pale glowing balm. Add vertical gold Chinese-style typography: a slim framed label on the far left reading 「掌中舞香膏」, the large main calligraphy 「赵飞燕」, smaller copy including 「回眸一笑百媚生」, and a right-side vertical phrase 「深宫秘香 惑人心魄」. Along the bottom place exactly 3 circular feature badges labeled 「古法合香」, 「名贵香料」, and 「长久留香」, plus a small fine-print line in gold.

Right panel: Create a moonlit palace-water fragrance poster titled {argument name="right main calligraphy" default="洛水有神 肌若凝露"}. The setting is a dark lotus pond or palace bath at night, with reflective water, distant windows, mist, carved architecture, and cold blue lighting. A young Chinese woman is seen from behind in a black translucent embroidered gown with an open back, wet-looking hair in an ornate updo with silver floral hairpins and dangling ornaments; her face is turned slightly sideways and obscured by a soft square blur. In the lower-left/center foreground place exactly 1 pale celadon jade-green glass perfume bottle with a round cap on a carved dark pedestal, with raised floral relief and small vertical product text on the bottle. Add elegant vertical calligraphy on the upper-left reading 「洛神凝露」 and large text 「洛水有神 肌若凝露」, with a small red seal stamp. Add narrow vertical copy on the right edge, a signature-like mark near the lower right, a centered bottom slogan reading approximately 「凝洛水之灵气 养倾城之美肌」 and 「洛神凝露 复刻古方 香养肌肤」. Along the bottom place exactly 4 circular icon badges, each with a simple line icon and short Chinese label, arranged horizontally.

Visual style: Ultra-detailed photorealistic fashion editorial mixed with Chinese historical fantasy, luxurious perfume packaging design, soft haze, volumetric light, cinematic shadows, shallow depth of field, ornate gold foil lettering, authentic ancient Chinese decorative motifs, refined poster layout, high contrast, premium cosmetics advertising.

Constraints: Keep exactly 2 poster panels, exactly 3 product objects in the left panel, exactly 1 product bottle in the right panel, exactly 3 bottom badges on the left, and exactly 4 bottom badges on the right. Preserve the Chinese characters as decorative text where specified, avoid modern objects, avoid extra panels, avoid cartoon style, no watermark, no English headline text.

Create a highly detailed photorealistic product-display scene of an adult anime-style collectible figure in a rainy neon Tokyo shopping street at night. The main subject is {argument name="character name" default="Yuki Miho"}, shown as a glossy 1/6 scale figurine squatting on a transparent circular acrylic stand in the center foreground, wearing a wet white one-piece swimsuit, long soaked dark brown twin-tail hair with small pink ties, and holding both hands up in double peace signs. Her face is deliberately obscured by a smooth square blur/censor block. A red leash or ribbon is attached near her neck and stretches diagonally from her collar toward the right foreground, creating a dramatic perspective line. The environment is a rainstorm with visible raindrops, wet reflective pavement, umbrellas, blurred pedestrians, and dense neon signs reminiscent of Shibuya or Akihabara; use cinematic depth of field, high contrast reflections, and ultra-real resin/plastic surface detail. Include exactly two price display elements: 1) a clear acrylic sign on the lower left reading 「結城 美柑」 and 「¥15,800」 with a red bottom stripe, and 2) a large product card or box panel on the right showing the same character in a beige school uniform jacket, green plaid skirt, teal bow, long twin-tail hair, with her face also blurred, plus the printed text 「結城 美柑」 and 「¥15,800」. Keep the scene wet, glossy, and commercial like a premium figure showcase photographed at street level; no extra characters in the foreground, no watermark, no additional price tags.

Goal: Create a clean Japanese presentation slide about AI as an approachable conversation partner, in a polished PowerPoint-style corporate/educational layout.

Canvas: 16:9 widescreen slide, white background with very subtle warm beige accents, thin black outer border, generous margins, minimalist and calm.

Header: Top-left small section title in Japanese: 「AIは気軽に聞ける相手」 with a short thin brown underline beneath it. Near the top center-left, add a pink circular badge containing 「4」 followed by pink annotation text: 「リポストしてるCodexのスキルで作ったパターン　ただのちゃんとした資料。すごすぎ。」 Top-right place a simple black JohoSauce-style logo: a droplet falling into ripple rings, with the word 「JohoSauce」 underneath.

Main title area: Centered large bold Japanese headline: 「{argument name="main headline" default="気軽に何でも聞ける相手"}」. Beneath it, smaller gray subtitle: 「AIは、怒らない・話し言葉OK・いつでも対応の相談相手です。」

Main content: Use exactly 3 equal rounded rectangular cards in one horizontal row, each with a thin light-gray border, white fill, soft shadow, and centered dark-brown line icon at the top. Card 1 icon: smiling chat bubble. Card 1 title: 「何度聞いても怒らない」. Card 1 body: 「同じ質問を何回しても大丈夫。\n3回でも5回でも、きちんと答えてくれます。」 Card 2 icon: side profile head speaking with two short sound lines. Card 2 title: 「話し言葉のままでOK」. Card 2 body: 「『えーっと、あの、なんていうか…』\nのような話し方でも汲み取ってくれます。」 Card 3 icon: simple analog clock. Card 3 title: 「24時間いつでも返ってくる」. Card 3 body: 「夜中の2時でも大丈夫。\nすぐに答えを返してくれます。」 Use exactly these three cards and no additional cards.

Bottom message bar: Add one wide pale beige rounded rectangle spanning most of the slide width. On the left inside it, place a simple outlined speech-bubble icon. To the right, bold Japanese message: 「{argument name="bottom message" default="構えず、まず話しかけてみる。それがAIと仲良くなる第一歩です。"}」

Visual style: Minimal Japanese business deck aesthetic, black and dark brown typography, pink accent only in the top annotation, simple monoline icons, lots of whitespace, balanced alignment, crisp vector-like rendering, no photos, no gradients except extremely subtle warm background shading, no extra decorative elements.

Create a highly detailed, glossy anime idol itasha scene at night: a cute pastel compact car resembling a rounded Volkswagen Beetle is parked three-quarter front view on wet reflective pavement under blooming cherry blossom trees, in front of a warmly lit Japanese shrine with a glowing sign reading {argument name="shrine sign text" default="ルミエール神社"}. The car is fully wrapped in a magical idol theme, with pearlescent pink, lavender, baby blue, and white paint, holographic sparkles, star decals, ribbon decals, lace-like trim, and neon purple underglow. Add exactly 8 prominent exterior decoration groups on the car: 1) a large roof ornament shaped like a pink winged star idol crest, 2) a hood graphic with bows, stars, and the English text {argument name="hood slogan" default="Memorial Idol"}, 3) a custom front license plate with tiny idol branding, 4) a pastel star badge on the side mirror, 5) a large printed anime idol portrait on the passenger door with the face area intentionally blank or softly obscured, 6) large cursive side text reading {argument name="side message" default="Thank You!"}, 7) glowing star wheel covers on both visible wheels, and 8) purple-pink LED strips along the bumper, side skirt, and wheel rims. Beside the car on the right stands a faceless young female idol mascot, full body, wearing a frilly pastel stage dress in blue, pink, and lavender with layered ruffles, bows, puff sleeves, white gloves, white platform ankle boots, and a small white hat; she holds a silver microphone in one hand and reaches toward the viewer with the other, with her face deliberately featureless/blurred for customization. Include exactly 1 dog at her feet, a small Cavalier King Charles Spaniel with brown-and-white fur, wearing a pastel bow tie, sitting calmly. The atmosphere is dreamy and celebratory, with falling sakura petals, soft bokeh lanterns, reflections in puddles, cinematic rim lighting, ultra-polished 3D illustration style, idol concert fantasy aesthetic, high saturation, crisp details, no extra people, no visible watermark. Use the idol theme {argument name="idol name" default="Kana"} and keep the overall mood affectionate, magical, and fan-made.

Goal: Create a cinematic storyboard sheet for a 15-second fantasy nature film about {argument name="main creature" default="a giant gentle ancient aquatic creature"} swimming through a mystical night wetland.

Canvas: Horizontal 16:9 image formatted as a storyboard contact sheet with exactly 12 panels arranged in a 3-column by 4-row grid. Use thin black gutters and subtle light gray panel borders. Add small white panel numbers 1 through 12 in the upper-left corner of each panel. No other text.

Visual style: Ultra-detailed cinematic concept art, dark blue and teal color palette, glowing bioluminescent forest, luminous vines, wet leaves, floating lotus leaves, reflective water surface, underwater caustics, bubbles, suspended particles, small glowing fish, volumetric moonbeams, dramatic depth, high contrast, dreamy photoreal fantasy atmosphere. The mood is awe and serenity, not horror or combat.

Panel content, exactly 12 discrete storyboard frames:
1. Split-level ultra-wide shot: the waterline cuts across the frame, above it a teal bioluminescent swamp forest, below it a deep underwater world; lotus leaves and ripples on the surface, with the huge creature’s shadow passing in the distance.
2. Low gliding water-surface shot: camera almost touching the reflective water, luminous hanging vines and wet foliage ahead, teal reflections streaking across ripples in the foreground.
3. Extreme macro close-up: a leaf tip in the upper-right corner holds a transparent water droplet glowing cyan; a second tiny droplet falls below it, with soft bokeh lights in the dark background.
4. Underwater splash moment: the droplet hits the water and sends bright circular ripples and bubbles downward; a school of tiny glowing fish scatters around the splash.
5. Wide underwater reveal: the giant aquatic creature appears far away as a graceful whale-like silhouette under blue shafts of light, surrounded by small fish and drifting particles.
6. Tracking shot beside underwater plants: tall swaying aquatic vegetation in the foreground, schools of glowing fish crossing the frame, and the massive creature’s body moving in the background to show scale.
7. Upward view from beneath submerged tree roots: enormous roots and a glowing tree base loom overhead, turquoise light streams down through water and hanging vines, creating a sacred forest-underwater composition.
8. Overhead shot of the water surface: circular ripples, clusters of lotus leaves, scattered blue sparkles, and the dark silhouette of the giant creature swimming below.
9. Half-submerged lateral shot: the upper half shows the glowing night forest and reflective swamp surface, the lower half shows the giant creature swimming just beneath the waterline.
10. Extreme close-up of the creature’s head and eye: gentle intelligent eye, whale-like skin covered in fine cyan bioluminescent speckles and flowing light patterns, soft bubbles around the face.
11. Climax action frame: the creature twists gracefully through the water, creating bright luminous trails, splashes, bubbles, and rippling caustic light beams from above.
12. Quiet final echo: return to a calm split-level wide shot of the glowing forest and dark underwater space, with the creature receding as a distant silhouette beneath the reflective surface.

Constraints: Keep the storyboard clean and readable as a single sheet, preserve exactly 12 numbered panels, maintain consistent teal-blue bioluminescent lighting across all panels, make the creature majestic and benevolent rather than scary, and avoid captions, speech bubbles, watermarks, logos, or extra labels.

Goal: Create a clean technical architecture diagram explaining {argument name="model name" default="YOLOv13"} as a one-page educational slide, showing how an input image flows through Backbone, HyperACE, Neck, FullPAD, and Head detection outputs.

Canvas: Wide 16:9 landscape infographic on a white background, crisp vector style, thin colored outlines, rounded rectangles, arrowed connectors, and readable sans-serif text. Use a left-to-right pipeline with supplementary modules below the main path.

Layout: Start at far left with a small stacked thumbnail illustration labeled “Input Image (3, H, W)”, then an arrow into a large blue Backbone panel. Place the green Neck panel in the upper center, the purple Head panel on the right, HyperACE in the lower left-center, FullPAD in the lower center-right, and a boxed explanatory caption along the bottom.

Main pipeline sections and exact element counts:
- 1 Input Image stack at far left, shown as layered landscape photos.
- 1 Backbone panel titled “Backbone” with subtitle “(built from DS-C3k2 blocks)”. Inside it, show exactly 5 vertically stacked DS-C3k2 block boxes labeled B1, B2, B3, B4, B5. Put “/2” downsampling markers between each adjacent block, for exactly 4 downsampling labels. Add side output arrows: B1 and B2 go right and are labeled “B1 (unused)” and “B2 (unused)”; B3, B4, and B5 go right in blue and are labeled “B3 (C3, H/8, W/8)”, “B4 (C4, H/16, W/16)”, and “B5 (C5, H/32, W/32)”.
- 1 HyperACE module below the Backbone, orange outline, titled “HyperACE” with subtitle “(Hypergraph-based Adaptive Correlation Enhancement) Eq.8-11”. Inside it, include one dashed rounded note box saying “Models cross-scale and cross-position high-order and low-order correlations”. Blue lines from B3/B4/B5 feed into HyperACE. HyperACE outputs an orange arrow to a small orange 3D cube labeled “Y” and “Correlation-enhanced feature”.
- 1 Neck panel in the upper center, green outline, titled “Neck” with subtitle “(built from DS-C3k2 blocks)”. Inside it show exactly 6 DS-C3k2 block boxes arranged in a 2-column by 3-row grid. Use black arrows between blocks, with exactly 2 yellow upsample boxes labeled “↑ 2×” and exactly 2 purple downsample boxes labeled “↓ 2×”. Include 3 blue circular input junctions on the left edge of the Neck and 3 green/purple circular tunnel markers below or beside the grid.
- 1 FullPAD legend module below the Neck, red outline, titled “FullPAD” with subtitle “(Full-Pipeline Aggregation-and-Distribution) Eq.12-13”. Show exactly 3 legend rows: “Backbone–Neck tunnel (Backbone-to-Neck connection)” using blue dotted arrows and blue hollow circles; “In-Neck tunnel (the internal layers of the Neck)” using green dotted arrows and green hollow circles; “Neck–Head tunnel (Neck-to-Head connection)” using purple dotted arrows and purple hollow circles. Connect the cube feature into FullPAD with an orange arrow, and draw dotted blue/green/purple tunnel routes upward and rightward to the Neck and Head.
- 1 Head panel on the right, purple outline, titled “Head” with subtitle “(Detect x3)”. Inside it show exactly 3 vertically stacked Detect boxes: “Detect (stride 8)”, “Detect (stride 16)”, and “Detect (stride 32)”. Each receives one arrow from the Neck and outputs to the right as H3, H4, H5 with labels “H3 (scale 1/8)”, “H4 (scale 1/16)”, and “H5 (scale 1/32)”.

Color coding: Backbone lines and panel in blue, Neck in green, Head in purple, HyperACE in orange, FullPAD outline in red. Main arrows are solid; FullPAD tunnels are dotted/dashed with hollow circular nodes. Use subtle pastel fills for module boxes.

Bottom caption: Add one wide rounded rectangle spanning the lower width. Text: “Classic YOLO is a one-directional Backbone → Neck → Head pipeline.” Then on the next line: “{argument name="model name" default="YOLOv13"} adds HyperACE and FullPAD on top of it to enhance feature representation and information flow.” Color the words Backbone blue, Neck green, Head purple, HyperACE orange, and FullPAD red.

Constraints: Keep all labels legible, align sections neatly, preserve the exact counts of blocks and detect heads, avoid extra modules, avoid photorealism except for the small input image thumbnail, and do not add a watermark.

Create a realistic, appetizing food photograph of Japanese omurice served on a large pale gray ceramic plate. The main subject is a glossy golden omelet draped over reddish ketchup fried rice, with the rice visibly peeking out along the bottom edge. On top of the omelet, draw a cute anime-style girl face using red ketchup lines: bobbed hair framing the face, short bangs, two large eyes, a small smiling mouth, and simple cheek/jaw lines, as if it is omurice art of {argument name="character name" default="a cute anime girl"}. Place a side salad on the right side of the plate with exactly 3 visible components: leafy green lettuce, dark red/purple leaves, and 1 cherry tomato, plus a light creamy dressing drizzle. Set the plate on a warm wooden café table near a softly lit window. Include exactly 2 background glass items: 1 small glass vase with tiny white flowers on the left and 1 clear water glass on the right, both softly blurred. Use natural morning sunlight, shallow depth of field, warm cozy tones, high-detail food texture, realistic reflections on the omelet, and a horizontal 16:9 composition. No text, no watermark, no hands, no utensils.

Goal: Create a dense Japanese reference poster/infographic titled {argument name="headline text" default="廊下・階段・通路演出の用語表"}, a vocabulary chart for illustrating hallways, stairs, and passageway scenes. Make it look like a hand-drawn scrapbook study sheet for AI illustration prompt words, with many miniature scene examples featuring the same mysterious gothic schoolgirl guide.

Canvas: Vertical 4:5 poster, warm ivory aged paper background with visible paper grain, thin ruled grid lines, taped corners, tiny stars, moons, flowers, stationery doodles, ink bottle, envelope, ribbon, and washi tape decorations around the border. Overall palette is sepia beige, muted navy, smoky gray, faded blue, and black ink. Use an analog watercolor, pencil, and ink style with slightly messy handwritten Japanese typography.

Main character: A recurring {argument name="character description" default="gloomy Jitome girl with long messy black hair, dark school uniform, pale skin, sleepy eyes, and a melancholic expression"}. She appears once large at the upper left as the guide and repeatedly inside the example thumbnails. The large portrait has a small label reading 「ジト目女子が案内するよ」. Keep the character’s face visible and consistent where she appears, with wet-looking hair strands and gothic, moody styling.

Header layout: At the top center place the large title 「廊下・階段・通路演出の用語表」. Beneath it add a subtitle line like 「学校廊下・病院廊下・非常階段・地下通路・長い廊下・消失点を整理した、AIイラスト発想補助のための通路演出ワード一覧。」 On the upper right add a decorative note card with a crescent moon and the text 「通路は物語の流れと心の流れをつくるよ。」

Central layout: Use exactly 24 numbered rectangular cards in a 4-column by 6-row grid. Each card must have a colored number tab, a Japanese category title, a few tiny handwritten keyword lines, and a small watercolor scene thumbnail. The thumbnails should show corridors, stairs, passageways, windows, vanishing points, reflections, silhouettes, or the gothic girl in different positions. Count and include these 24 cards exactly:
1. 「学校廊下」 — school hallway with lockers, bulletin boards, afternoon light.
2. 「病院廊下」 — quiet hospital corridor with white walls, curtains, clean unease.
3. 「非常階段」 — emergency stairwell, metal rails, cold light, escape-route tension.
4. 「商業施設通路」 — shopping mall passage with people flow and urban brightness.
5. 「地下通路」 — low ceiling, artificial lighting, tiled walls, damp heaviness.
6. 「長い廊下」 — deep vanishing point, repeating windows and doors.
7. 「手すり」 — railing lines guiding the eye along stairs or corridor.
8. 「消失点」 — one-point perspective corridor converging into distance.
9. 「階段」 — up/down movement, diagonal composition, a person midway.
10. 「踊り場」 — stair landing, pause between floors, quiet turning point.
11. 「外階段」 — outdoor stairs with fence, sky, wind, exposed metal.
12. 「連絡通路」 — connecting passage between buildings, movement and transition.
13. 「渡り廊下」 — covered school walkway with view outside and nostalgia.
14. 「狭い通路」 — narrow passage, close walls, body brushing through.
15. 「薄暗い通路」 — dark corridor, weak light, humid uneasy atmosphere.
16. 「明るい廊下」 — open bright hallway, clean floor, hopeful transparent light.
17. 「窓のある廊下」 — windows, curtains, light and shadow bands.
18. 「曲がり角」 — corner turn, unseen continuation, narrative suspense.
19. 「通路の奥」 — far end of corridor, small figure, depth and destination.
20. 「すれ違い」 — two characters passing each other without eye contact.
21. 「立ち止まる」 — character stopped in passage, back view, quiet hesitation.
22. 「振り返る」 — character looking back over shoulder, hair and skirt movement.
23. 「追いかける背中」 — chasing or following someone’s back in a passage.
24. 「通路の余韻」 — empty corridor after people leave, lingering atmosphere.

Bottom section: Add a handwritten notes area with star bullet points explaining that this is a word list for school corridors, hospitals, stairs, and passage演出, useful for combining with keywords like long hallway, staircase, railing, vanishing point, curve, light, shadow, silhouettes, and passing-by scenes. Keep the text dense but legible, like a Japanese prompt-word cheat sheet.

Visual style: Delicate anime watercolor illustration, sketchbook infographic, hand-lettered Japanese, thin ink outlines, soft blue-gray shadows, warm parchment paper, collage tape, decorative celestial motifs. The poster should feel information-dense, cozy, nostalgic, slightly gothic, and handmade.

Constraints: Use exactly 24 numbered cards in the grid; keep the large title readable; avoid modern flat vector style; no photorealism; no English labels except decorative tiny marks if needed; no watermark or logo.

Goal: Create a vertical American comic book cover in a gritty punk-rock manga/anime style for a fictional issue titled {argument name="comic title" default="HEARTSTRIKE"}.

Canvas: Portrait comic cover, 2:3 aspect ratio, full-bleed illustration with a black distressed print border, halftone texture, ink speckles, worn paper scratches, high-contrast magenta, black, white, and purple lighting.

Main subject: Center a young female punk rock guitarist idol in dynamic low-angle perspective, filling most of the cover. She has blonde twin-tail hair, pointed elf ears, a red plaid beret with a small heart pin, black ribbon hair ties, gothic punk makeup vibe, a black corset top with frills, fishnet/mesh sleeves, studded bracelets, choker, red plaid skirt, belts, chains, lace trim, and a heart-shaped garter detail. Her face is deliberately covered by a plain skin-tone rectangular censor block with no facial features. She is aggressively playing a white Stratocaster-style electric guitar angled diagonally from lower left to mid-right, with detailed strings, pickups, knobs, worn scratches, and a visible headstock signature.

Background: A chaotic concert stage packed with silhouetted fans raising rock hand signs, purple lightning bolts, starburst stage lights, smoke, electric energy, and black speaker amps on the right. The mood is loud, rebellious, dangerous, and glamorous.

Cover typography and layout: Use exaggerated distressed comic lettering with jagged edges, pink shadows, white fills, black outlines, and ink splatter. Include exactly 8 visible text blocks: 1) top-left publisher box: "REBEL INK COMICS" with a pink skull-and-bow emblem; 2) price/issue box: "#1", "$4.99", "US", "APR", "2025"; 3) top tagline across the cover: "SHE PLAYS LOUD. SHE HITS HARD. SHE NEVER MISSES." with the words "LOUD", "HARD", and "NEVER" in hot pink; 4) huge main masthead: "{argument name="comic title" default="HEARTSTRIKE"}"; 5) left promo stack: "1ST SMASH ISSUE!"; 6) left slogan: "LIVE. LOVE. DESTROY." with "DESTROY." in hot pink; 7) right tilted caption: "FEATURING: THE WORLD'S MOST DANGEROUS SOUND!" with "DANGEROUS" in hot pink; 8) bottom-right hot pink tour stamp: "{argument name="tour stamp text" default="BLACKOUT TOUR '25 BEGINS!"}". Add a small vertical barcode box at bottom left with "DIRECT SALES" and barcode numbers.

Visual style: Fusion of American punk comic cover, Japanese anime character rendering, heavy cel shading, detailed line art, distressed screenprint texture, neon magenta rim light, purple concert glow, dramatic foreshortening, sharp inked outlines, glossy highlights on guitar hardware, chaotic composition.

Constraints: Keep the cover busy but readable, preserve the exact count of 8 major text blocks, do not add extra title banners or logos beyond the listed elements, keep the face fully hidden by the plain rectangle, and make it look like a collectible first-issue comic cover.

Using REFERENCE_0 as the base for the blonde school-uniform girl, transform it into a realistic cinematic scene from a Japanese classroom. Show the girl now standing indoors, facing a male classmate, with an embarrassed shocked expression and open mouth. Keep her blazer, cardigan, bow tie, plaid skirt, blonde curled hair, and overall fashion-leader schoolgirl identity recognizable, but add rural farmer styling over the uniform: a white floral headscarf tied around her hair, loose dark floral work pants worn under the skirt, and muddy rubber boots. Add exactly 1 other person: a Japanese male student seen mostly from behind/side in a navy school blazer and plaid trousers, standing on the left and facing her. Place them in an old classroom with wooden floors, rows of desks and chairs, a green chalkboard, classroom posters, tall windows, and warm late-afternoon sunlight. Use a vertical full-body composition, photorealistic live-action look, natural lens depth, dramatic novel-scene atmosphere, no captions or extra text.

Create a vertical high-fashion fantasy portrait of {argument name="character name" default="an anonymous woman"} from the waist up, posed in three-quarter profile facing left, with her face deliberately covered by a centered opaque peach rectangular block for anonymity. She has voluminous flowing {argument name="hair color" default="strawberry-blonde and copper-red"} hair whipped upward and outward as if caught in wind, merging with enormous painterly floral and fabric forms. Surround her head and body with exactly 6 prominent flower elements: 2 large red hibiscus-like blossoms above and left of the head, 1 large teal-blue flower on the right side, 1 pink daisy near the cheek area, 1 blue flower at the waist, and 1 small pale flower near the bodice. Dress her in an ornate strapless couture gown covered in red, white, blue, and orange petals, sequins, lace, and splattered paint textures, with a jeweled multicolor necklace at the neckline. Add dramatic translucent ribbons and paint-like splashes in vivid crimson, vermilion, coral, orange, teal, cobalt, magenta, and white swirling around her, creating an explosive floral headdress and abstract background. Use a clean white studio backdrop, bright high-key lighting, shallow depth of field, glossy editorial photography realism mixed with surreal botanical couture, extremely detailed textures, dynamic motion, saturated colors, and a vertical 9:16 composition. Keep the face block plain and flat with no facial features or text, avoid logos or watermarks.

Goal: Create a polished Apple-style Korean infographic slide summarizing {argument name="headline text" default="WWDC 2026 HIGHLIGHTS"} from a designer perspective.

Canvas: 16:9 horizontal presentation graphic, 1152×768 feel, white-to-pale-blue gradient background with soft translucent wave ribbons in the top-right corner. Use Apple-inspired glassmorphism, rounded cards, subtle shadows, blue/purple gradients, clean SF Pro-like typography, and a small multicolor Apple logo at the top-left.

Layout: Large centered headline at the top, with “WWDC 2026” in dark navy and “HIGHLIGHTS” in blue-to-purple gradient. Under it, arrange exactly 6 rounded white cards in a 3-column by 2-row grid, each card numbered with a glossy blue-purple circular badge from 1 to 6.

Card 1: Title {argument name="liquid glass title" default="리퀴드 글래스 정교화"}. Show a translucent frosted-glass control panel with exactly 3 horizontal sliders: a toggle slider, a brightness slider with sun icon, and a volume slider with speaker icon. To the right, show a depth-enhancement slider and a stack of translucent glass sheets. Include exactly 3 Korean bullet points: “정밀한 투명도 조절로 더욱 몰입감 있는 경험”, “동적 조명과 섬세한 블러로 깊이감 향상”, “모든 플랫폼에 일관된 리퀴드 글래스 경험 제공”.

Card 2: Title {argument name="design principles title" default="디자인 원칙"}. Subtitle “AI 시대를 위한 7가지 핵심 원칙”. Show exactly 7 circular principle icons in a neat grid, labeled in English: Purpose, Agency, Responsibility, Familiarity, Flexibility, Simplicity, Craft, Delight. Even though 8 label positions are visible, emphasize the subtitle says 7 principles while the visual grid contains these 8 labeled icon circles. Bottom pill caption: “AI 윤리와 책임을 최우선으로 하는 디자인”.

Card 3: Title {argument name="ios title" default="iOS 27"}. Show a realistic iPhone mockup with dynamic island, blue-purple Siri orb wallpaper, and Korean text on screen “Siri가 듣고 있어요...”. To the right, include exactly 3 vertical feature rows with icons and Korean headings: “정리” with smaller text “알림과 앱을 자동으로 정리”, “확장” with smaller text “필요에 맞게 콘텐츠 확장”, and “재구성” with smaller text “내용을 이해하고 재구성하여 제안”. Bottom rounded pill: “더 스마트해진 Siri와 개인화된 경험”.

Card 4: Title {argument name="safari home title" default="사파리 & 홈"}. Show a Safari-like AI tab group module labeled “AI 탭 그룹” with subtitle “AI가 관련 탭을 자동으로 그룹화”. Include exactly 4 visible tab chips plus one plus button: “여행 계획 7”, “리서치 5”, “쇼핑 4”, and a final add button “+”. Below, show a Home security clipping section titled “홈 보안 클립” with subtitle “카메라가 감지한 주요 순간을 자동으로 요약”. Include exactly 4 video thumbnails with time labels: 9:12 AM delivery person, 1:47 PM car, 4:32 PM dog, 8:11 PM living room. Bottom pill: “AI가 더 안전하고 편리한 경험을 제공합니다”.

Card 5: Title {argument name="macos title" default="macOS 골든 게이트"}. Show an iMac display with Golden Gate Bridge sunset wallpaper and a glowing Apple Silicon chip badge. Under it, include exactly 3 benefit icons with Korean headings and short subtext: “성능 향상”, “배터리 효율”, “강력한 보안”. Bottom pill: “Apple Silicon의 힘으로 더 강력해진 Mac 경험”.

Card 6: Title “visionOS”. Create a futuristic spatial-computing scene: a glowing central blue-purple orb labeled “Siri Pro” hovering over a faint perspective grid. Around it place exactly 6 floating translucent app tiles/icons: image/photo icon, play icon, chart icon, person icon, profile silhouette icon, and one small glass panel. Bottom pill: “3D 공간에서 더욱 자연스럽고 지능적인 상호작용”.

Visual style: High-end Apple keynote recap aesthetic, crisp vector UI mixed with soft 3D glass elements, blue/purple glow accents, consistent spacing, rounded 24px cards, light shadows, premium minimalism. Keep all Korean text legible and correctly spelled. Avoid extra logos, watermarks, signatures, or additional cards.

Using the provided reference image, enhance and restore the same photo into an ultra-premium cinematic 8K-quality image. Preserve 100% of the original identity, pose, composition, framing, perspective, and all objects exactly as they are. Recover realistic microdetails and sharpness throughout: crisp eyes, detailed nose texture, individual fur strands, clear flower petals and stems, defined grass, house, trees, vehicle, and background elements. Make it photorealistic with natural sunlight, high contrast, studio-quality clarity, accurate colors, no pixelation, no blur, no artifacts, and no invented changes.