GPT Image 2 prompts

Create a premium, modern high-fashion CHARACTER CONCEPT ART SHEET in a 16:9 widescreen layout on a pure white background. 

THE ATTACHED REFERENCE IMAGE DEFINES THE ART STYLE — treat it as a strict style bible and reproduce its exact rendering

Create a single vertical anime animation development board for an original emotional short film titled "{argument name="film title" default="The Boy and the Umbrella"}".
The output must be ONE combined image containing an anime character design sheet and a cinematic storyboard page.

Create a playful full-body studio composite of {argument name="character name" default="a stylish teenage boy"} being followed by a tiny chibi version of himself. The main character is a realistic young man with voluminous dark brown wavy hair, clean face, relaxed slightly aloof expression, looking back over his shoulder while walking to the right. He wears a burgundy bomber jacket over a black hoodie, oversized faded blue cargo jeans with large side pockets, and gray-and-white sneakers; his right hand holds a clear plastic cup of iced tea or coffee. Behind him at lower left, include exactly 1 miniature chibi duplicate, about one third his height, with the same hair and outfit, exaggerated angry face, furrowed eyebrows, clenched teeth, and both hands gripping and pulling the loose fabric of the main character’s left pant leg backward, stretching the denim dramatically. Use a clean light gray seamless studio background with soft floor shadow, high-detail photorealistic rendering for the main teen, glossy 3D cartoon/anime styling for the chibi, and a humorous fashion-editorial mood. Add exactly 10 simple black hand-drawn doodle accents around the figures: 2 sparkle icons near the upper body, 1 sparkle icon to the right, 1 squiggly surprise mark near the main character’s head, 2 small steam-puff frustration clouds near the chibi, 1 anger symbol near the chibi’s head, and 3 short motion-line clusters around the stretched pant leg and feet. Keep the composition vertical, centered, with ample negative space, no text, no watermark, no extra characters.

Create a cinematic, polished, naturally realistic studio portrait of a {argument name="subject" default="handsome young adult man"} facing the camera head-on from mid-chest up. He has {argument name="hair color and style" default="thick tousled dark brown hair with volume"}, warm brown eyes, strong brows, subtle stubble and a faint mustache, a defined jawline, and a calm serious expression with closed lips. Dress him in a {argument name="clothing" default="plain off-white crewneck sweatshirt"}. Use dramatic movie-style lighting with a warm amber key/rim light from the viewer’s left catching the hair, cheek, ear, and shoulder, balanced by a cool cyan-blue rim/fill light from the viewer’s right. Set him against a dark smoky blue-black studio background with a soft teal glow on the right side, shallow depth of field, crisp facial detail, realistic skin texture, natural pores, glossy catchlights, and high-end editorial color grading. Center the subject symmetrically, vertical 4:5 composition, no text, no accessories, no watermark, no extra people.

Create a realistic intimate lifestyle portrait of {argument name="subject" default="a young adult woman"} lounging on her stomach on a dark charcoal sofa in a dim modern apartment at night, holding and looking over a smartphone near the bottom foreground. She has {argument name="hair color and style" default="dark brown shoulder-length straight hair swept to one side"}, fair skin, a calm slightly dreamy expression, glossy natural lips, and thin round metal eyeglasses with subtle green-blue phone reflections on the lenses. She wears {argument name="outfit" default="a cream silk camisole with lace trim and thin straps, with matching light shorts"}. Pose her diagonally across the couch with elbows forward, hands around the phone, shoulders and upper torso closest to the camera, one bent leg visible behind her. Use warm low ambient apartment lighting mixed with cool phone glow on her face and fingers, shallow depth of field, soft cinematic bokeh, realistic skin texture, natural proportions, and a candid indoor photography feel. Background: blurred beige walls, a dark hallway or kitchen entrance on the left, a small wall thermostat/switch on the back wall, and minimal furniture, all unobtrusive. Composition: vertical 2:3 portrait, eye-level slightly above the couch, 35mm lens look, close focus on face and hands, no text, no watermark, tasteful fashion/lifestyle mood, not explicit.

Goal: Create a dark futuristic widescreen tech poster introducing {argument name="headline text" default="今日分享：5 个 GitHub 爆涨星技能"}, showing five fast-growing GitHub AI tools as large interface cards above a modern developer workstation.

Canvas: 16:9 horizontal social media banner, cinematic dark office background, cool blue lighting, high contrast, sharp UI mockup style. Use a blurred glass-and-metal workspace at night, with a desk across the bottom third.

Layout: Place a large centered Chinese headline at the top; make the number “5” oversized and bright electric blue while the rest of the headline is bold white. Under it, add a thin decorative line and small subtitle text reading {argument name="subtitle text" default="真实工作流场景一图看懂"}. Arrange exactly 5 numbered feature sections in a row across the middle, each with a circular number badge, tool name, short Chinese description, and a large screenshot-style UI card beneath it. The five sections are: 1) GPT-Image2-Skill, description “图像生成与创意设计”; 2) smart-ralph, description “智能开发规划与执行”; 3) OpenAI Skills, description “可复用的 AI 技能目录”; 4) open-saas, description “开箱即用的 SaaS 基础设施”; 5) awesome-copilot, description “Copilot 配置与提示词库”.

Subject details: Section 1 card is a dark prompt-library interface titled “GPT-Image2-Skill,” with a left sidebar of categories and a grid of poster/UI/illustration thumbnails. Section 2 card is a dark coding workflow interface titled “smart-ralph,” showing a task checklist, code editor, checkpoints, and an approval dialog. Section 3 card is a clean light “Skills Directory” interface with category navigation and stacked skill cards such as Code Reviewer, API Documenter, SQL Assistant, and Data Analyst. Section 4 card is an open-saas dashboard mockup with a landing page headline “Build Faster. Launch Sooner.”, login panel, dashboard metrics, payments table, and deploy panel. Section 5 card is a dark awesome-copilot marketplace/library interface with tabs and exactly 6 visible tool rows: Python Best Practices, API Development, Test Generator, Code Refactor, Security Scanner, and Docs Generator.

Desk foreground: Include exactly 4 main desk objects: an open laptop on the left showing a mountain wallpaper, two wide desktop monitors in the center showing code and analytics dashboards, a white keyboard with mouse, and a small potted plant near the right monitor. Keep the desk minimal, premium, and realistic.

Visual style: Premium GitHub/AI tooling aesthetic, dark navy and black palette, electric blue accents, crisp white typography, glassmorphism UI panels, realistic reflections, subtle depth of field, cinematic lighting, polished startup presentation design.

Constraints: Keep all five tool names readable. Do not add extra tools or extra numbered sections. Preserve the Chinese headline and short Chinese descriptions. Avoid watermarks, logos unrelated to the five tools, clutter, distorted text, or cartoon styling.

Goal: Create a cinematic first-person shooter screenshot in Shanghai at night, blending gritty realistic military-game visuals with a surreal giant anime maid looming over the skyline.

Canvas: Wide 16:9 frame, dark rainy night, high-contrast blue-gray color grading, wet lens grime, film grain, subtle scratches, dramatic storm clouds, reflections on the river.

Scene and layout: The viewer stands on a riverside battlefield facing the Pudong skyline across dark water. The Oriental Pearl Tower is clearly visible on the far left, with illuminated skyscrapers and neon signs along the waterfront. In the lower right foreground, show a realistic black scoped assault rifle held in first-person view, rain-speckled and aimed toward the skyline. Add fires, smoke plumes, small explosions, and city lights reflected in the water. A helicopter with a bright searchlight appears in the upper right sky.

Main subject: A colossal semi-transparent anime giantess appears behind and above the skyscrapers, scaled so her head and shoulders dominate the sky. She is {argument name="character name" default="an original anime maid giantess"} with {argument name="hair color" default="pink-purple twin-tail hair with blue streaks"}, large purple eyes, small open mouth, and a mischievous or threatening expression. She wears a frilly maid-style headpiece and sailor-maid outfit with dark blue ribbons. Her left hand reaches forward toward the viewer in extreme foreshortening, with the palm and fingers huge in the foreground, as if trying to grab the city or the player. Make her softly luminous and slightly transparent, like a massive holographic or supernatural apparition in the rain.

HUD and readable overlays: Include exactly 5 distinct game-interface elements: 1 square minimap in the lower left, 1 centered crosshair/weapon status cluster near the bottom, 1 ammo counter panel in the lower right showing large numbers similar to “21 / 84”, 1 small status label above the ammo panel, and 1 row of small icons at the top right. Include exactly 2 centered Chinese subtitle lines near the bottom: {argument name="first subtitle line" default="汉娜：大家快看天上！"} and {argument name="second subtitle line" default="爱尔兰佬：妈的，这是啥玩意！"}. Add a small Chinese objective label near the minimap reading {argument name="objective label" default="抵达撤离点"}.

Counted visible elements to include: exactly 1 giant anime maid, exactly 1 oversized reaching hand, exactly 1 helicopter with searchlight, exactly 1 first-person scoped rifle, exactly 1 Oriental Pearl Tower, exactly 2 subtitle lines, and exactly 5 HUD/interface groups. Do not add extra characters in the foreground.

Visual style: Photorealistic AAA military FPS screenshot with realistic rain, wet surfaces, smoky battlefield atmosphere, cinematic lighting, and authentic UI overlays; contrast the realistic war scene with a crisp anime character rendered as a gigantic translucent figure.

Goal: Create a colorful Japanese manga-style food review comic about a chibi character visiting a Kobe craft beer restaurant and enjoying small-size tasting portions.

Canvas: Vertical 3:4 comic page, bold black panel borders, bright saturated colors, clean digital manga line art, halftone dots and speed-line backgrounds, expressive chibi proportions. The main character is {argument name="character name" default="a chibi black-haired boy with rectangular glasses"}, wearing {argument name="outfit" default="a red beret, orange T-shirt, blue jeans, and black sneakers"}. Use thick outlines, exaggerated facial expressions, and readable Japanese speech bubbles.

Layout: Use exactly 5 comic panels.
1. Top-left panel: The character sits at a wooden restaurant table studying a menu and writing with a pen, looking serious and slightly confused. Add one oval speech bubble reading 「種類が多い…」.
2. Top-right panel: The character walks up a narrow exterior stairway beside a standing yellow signboard that reads “KOBE BEER” with small Japanese text beneath it. The character points excitedly. Add two speech bubbles reading 「神戸の地ビールかー」 and 「珍しいな」.
3. Large center panel spanning the width: The character sits at a wooden table eating from a bowl with a spoon, smiling widely. Show exactly 5 food and drink items on the table: a red craft beer glass labeled “Craft Kobe,” a golden craft beer glass labeled “Craft Kobe,” a bowl of soup or stew, a plate of sliced prosciutto, a small salad plate, and a plate of fried chicken karaage. Add two jagged manga burst speech bubbles reading 「いろいろ楽しむぜ！」 and 「Sサイズで…」. Use a warm yellow radial burst background.
4. Bottom-left panel: Close-up of the happy character laughing with a pink speed-line background. Add two speech bubbles, one oval reading 「唐揚ぎもうまい」 and one jagged reading 「うまーい！ぷはぁ〜」.
5. Bottom-right panel: The character nervously holds a large golden beer glass with foam, sweating slightly, seated at the table with a dark blue uneasy background and wavy stress lines. No extra dialogue except small sweat/emotion marks.

Text content: Preserve all Japanese text exactly as written, including 「種類が多い…」, 「神戸の地ビールかー」, 「珍しいな」, 「いろいろ楽しむぜ！」, 「Sサイズで…」, 「唐揚ぎもうまい」, and 「うまーい！ぷはぁ〜」. The beer sign must say {argument name="beer sign text" default="KOBE BEER"}. The glass labels must say {argument name="glass label text" default="Craft Kobe"}.

Visual style: Playful shonen manga restaurant report, crisp inked outlines, white speech balloons with black Japanese lettering, energetic panel composition, appetizing food illustration, cozy wooden tabletop details, lively but not photorealistic.

Constraints: Use exactly 5 panels, exactly 7 Japanese speech bubbles across the comic, and the listed visible food/drink arrangement. Do not add extra menu writing beyond small unreadable marks, do not add additional characters, and avoid photorealism.

Create a photorealistic 3D studio product render showing exactly three upright infinity-loop sculptures shaped like the number 8, arranged side by side on a glossy reflective gray floor. The left sculpture is transparent clear glass with thick rounded edges, visible refraction, caustic highlights, and subtle internal reflections. The center sculpture is polished mirror chrome metal, highly reflective with dark and bright warped studio reflections across its smooth twisted surface. The right sculpture is white marble with soft gray veining, matte-satin stone texture, and rounded carved edges. All three objects have identical continuous Möbius-like figure-eight forms with an open oval hole at the top and bottom, standing vertically and casting soft shadows plus faint floor reflections. Use a minimalist warm-gray studio background with blurred rectangular softbox light reflections on the left and right walls, diffuse natural lighting, shallow depth of field, centered symmetrical composition, high-end luxury material study, no text, no people, no watermark. Use {argument name="object count" default="three"} sculptures, {argument name="left material" default="transparent clear glass"}, {argument name="center material" default="polished mirror chrome metal"}, {argument name="right material" default="white marble with gray veining"}, and {argument name="background color" default="warm gray"}.

Goal: Create a colorful Japanese manga page about a chibi character enjoying Kobe beer and food, in a bold anime comic style with thick black outlines, expressive speed lines, halftone textures, and clean digital coloring.

Canvas: Vertical 3:4 manga page, approximately 768 x 1024, divided by thick black gutters into exactly 5 panels: 2 small panels across the top row, 1 large full-width centerpiece panel in the middle, and 2 panels across the bottom row.

Character details: Use one recurring chibi manga character in all 5 panels: a small youthful person with spiky black hair, rectangular black glasses, a bright red beret tilted backward, an orange T-shirt, expressive eyebrows, and exaggerated open-mouth reactions. Keep the character consistent across panels.

Panel count and layout:
1. Top-left panel: the character sits at a wooden table studying a menu or checklist with a pen, looking serious and slightly overwhelmed. Add a small sigh puff near the face and one oval speech bubble containing {argument name="top left bubble text" default="種類が多い…"}.
2. Top-right panel: the character runs excitedly through a narrow street or alley with stairs in the background. Place a yellow sidewalk sign reading “KOBE BEER” in large black letters. Include two speech bubbles: one vertical bubble containing {argument name="top right large bubble text" default="神戸の地ビールかー"} and one smaller bubble containing “珍しいな”.
3. Center large panel: the character sits at a table joyfully eating ramen with chopsticks, pulling up long noodles from a steaming bowl, smiling wide with sweat drops of excitement. Around the bowl are exactly 4 visible food/drink groups: a red craft beer glass labeled “Craft Kobe”, a golden craft beer glass labeled “Craft Kobe”, a tall dark beer glass, and three food plates consisting of sliced meat, a small salad, and fried chicken pieces. Use a radiant yellow burst background. Add two jagged speech balloons: left balloon contains {argument name="center left bubble text" default="いろいろ楽しむぜ！"}; right balloon contains {argument name="center right bubble text" default="Sサイズで…"}.
4. Bottom-left panel: close-up of the character grinning with mouth wide open, satisfied and excited, with pink speed lines in the background. Add two speech bubbles: one oval bubble containing “唐揚ぎもうまい” and one jagged bubble containing {argument name="bottom left jagged bubble text" default="うまーい！ぷはぁ〜"}.
5. Bottom-right panel: the character drinks a large foamy beer from a tilted glass with both hands, cheeks slightly puffed, sweat drop on the face, blue vertical speed-line background, wooden tabletop at the bottom. Show splashes or foam droplets near the glass.

Visual style: Cute chibi Japanese comic, high-energy food manga expression, bold black panel borders, crisp line art, saturated colors, playful exaggerated typography in speech bubbles, no realism, no watermark, no extra panels, no extra characters.

Using REFERENCE_0 as the source material, create a dark-mode Japanese evaluation report slide that compares attempts to generate a character sheet while preserving the same rough watercolor/ink illustration style.

Goal: Turn the reference into a structured experiment-summary dashboard titled 「M2　ステップ2—ムードボードを見本に GPT でキャラクターシート」. Keep the subject as the small armored child character from the reference, but present the result as an analysis screen rather than a single illustration.

Canvas and style: Use a wide 16:9 dark UI layout with charcoal background, thin dividers, rounded cards, small Japanese UI text, orange/green accent labels, and embedded white image panels. Make it look like a screenshot of an internal comparison report.

Top section: Create exactly 3 comparison cards across the top. Each card contains a white character-sheet panel showing the same character in multiple views and a small detached head/bust crop. Use these card titles and edge scores:
1. 「経路A：GPT ムードボード経由」 with 「エッジ 57.6」
2. 「経路B：grok ムードボード経由」 with 「エッジ 52.6」
3. 「参考：grok キャラ直接（B-5）」 with 「エッジ 48.8」
Under each card add one short Japanese note line; the third card should mention that it is the softest and inherits the sample’s green cloth.

Middle section: Add a comparison table with exactly 4 columns labeled 「条件」, 「エッジ強度」, 「コントラスト」, 「見本の欠陥」. Include exactly 7 data rows:
1. 「grok（目標）」 — 47.2 — 82.4 — 「—」
2. 「1) 共通のみ」 — 53.6 — 90.8 — 「—」
3. 「2) + 筆致（言葉）」 — 52.0 — 88.4 — 「なし」
4. 「3) + grok キャラ見本」 — 「48.8 ← 最良」 — 85.4 — 「緑の布を継承」
5. 「4) + GPT ムードボード」 — 「57.6 ← 最悪」 — 84.5 — 「なし」
6. 「5) + grok ムードボード」 — 52.6 — 87.3 — 「なし」
7. 「素材 image3.png（参考）」 — 58.8 — 69.1 — 「—」

Bottom section: Add several short Japanese conclusion lines beginning with 「結論1」 through 「結論5」, explaining that moodboard routes are not very effective, direct character reference works better, and the practical workaround is to avoid moodboards and use the Grok output with cropping. Finish with a collapsible dark bar labeled 「プロンプト（ステップ2の冒頭：ムードボードの役割指定）」.

Constraints: Preserve the character identity and rough painted look from REFERENCE_0 in the embedded panels, but do not recreate the full original reference screenshot. Do not add unrelated characters. Use crisp UI typography and readable Japanese labels. The main title topic may be customized as {argument name="report title" default="M2　ステップ2—ムードボードを見本に GPT でキャラクターシート"}; the three route labels may be customized as {argument name="route labels" default="経路A：GPT ムードボード経由 / 経路B：grok ムードボード経由 / 参考：grok キャラ直接（B-5）"}; the character-sheet subject may be customized as {argument name="character subject" default="the small armored child knight from the reference"}; the table metric emphasis may be customized as {argument name="best and worst edge notes" default="48.8 ← 最良, 57.6 ← 最悪"}.

Create a dynamic high-detail anime battle illustration showing {argument name="attacker name" default="Origin-sama"} violently slamming {argument name="victim name" default="a pink-haired magical girl"} headfirst into a cracked stone floor. The attacker is a fierce young woman with split black-and-white hair accented by vivid red streaks, long thick twin braids whipping outward like ropes, small round hair buns, sharp teal eyes, and a black star-pattern mask across her eyes. She wears a black tailored jacket over a white ruffled blouse, a short black skirt with red lining and white frills, black shoes, and a dramatic combat pose with one knee raised while her hand grips the victim’s head. The victim has very long flowing pastel pink hair, a frilly white magical-girl dress with orange trim and ribbons, white gloves, and is shown upside down at the moment of impact, one hand braced on the rubble. Emphasize the hit-reaction scene rather than a clean attack pose: the victim’s hair explodes across the foreground, the ground shatters into a crater, rocks and dust fly in all directions, and bright pink-red energy ribbons spiral around both characters. Use an intense low-angle vertical composition, cinematic foreshortening, motion blur, debris arcs, dramatic lighting, dark smoky background, glowing highlights, and polished modern anime rendering with extremely detailed fabric, hair strands, impact particles, and swirling shockwave effects. No text, no speech bubbles, no watermark.

Create a cinematic photorealistic fine-dining scene in a warm modern kitchen at blue hour: an Asian male chef in a dark apron stands on the left side of the frame, partially cropped, concentrating as he uses tweezers and chopsticks to place tiny herb garnishes on an elegant plated dish. The dish sits centered on a large dark stone kitchen island in a shallow wide ceramic bowl, featuring a glossy seared fish fillet with golden char marks, small yellow vegetable batons, red and green accents, edible flowers, shiso-like leaves, and a small mound of black rice or caviar-like garnish. The kitchen has dark wood cabinetry, black stone backsplash, open shelves with bowls and plates, hanging towels, cooking utensils, bottles, and warm under-cabinet lighting. On the right, include a large panoramic window showing a twilight coastal forest cliff and deep blue ocean/sky outside, with three visible black pendant lights glowing warmly near the window. Use dramatic cinematic lighting: amber interior practical lights contrasting with cool blue dusk outside, soft shadows, realistic reflections on the countertop, shallow depth of field, high dynamic range, rich textures, natural skin detail, and an artful editorial food-photography composition. No text, no watermark, no extra people.

Create a photorealistic outdoor portrait of a {argument name="subject" default="young adult East Asian woman"} crouching on a sunlit rooftop terrace in a quiet hillside town. She has long straight dark brown hair with soft bangs, fair skin, natural makeup, and a calm neutral expression while looking directly into the camera. She wears a slightly loose pale blue-white button-up dress shirt with rolled sleeves, a loosened black necktie, a black skirt, and glossy black pointed high heels. Pose her low to the ground with knees bent apart, elbows relaxed forward, hands hanging naturally between her legs, balanced on her heels. The setting is a white concrete rooftop parapet with a gray tiled floor; behind her is a softly blurred village of traditional rooftops, utility wires on the left, green forested mountains, and a bright pale blue sky. Use natural midday sunlight, crisp highlights on the face and clothing, soft shadows on the floor, shallow depth of field, realistic skin texture, high-resolution DSLR look, 50mm portrait lens feel, centered composition, vertical 3:4 framing. Avoid text, logos, watermarks, extra people, exaggerated anatomy, or overly stylized rendering.

Create a serene, photorealistic vertical image of {argument name="subject" default="a young adult woman"} sitting sideways on a simple wooden chair on a sunlit balcony, reading an open paperback book and gently looking back toward the viewer. She has {argument name="hair color and style" default="long dark brown softly wavy hair with bangs"} and wears a minimal {argument name="outfit" default="warm beige two-piece swimsuit"}. Frame the scene from inside a traditional Japanese-style room or doorway, with wooden sliding door frames and translucent paper panels on the left and a warm cream stucco wall on the right, creating a natural architectural frame around the subject. The balcony overlooks {argument name="view" default="a quiet rocky inlet with clear turquoise seawater, a small shrine-like wooden structure on the shore, pine trees, green hills, and the open sea in the distance"}. Use warm late-afternoon sunlight, soft natural shadows, a peaceful vacation atmosphere, realistic skin tones, shallow depth of field with the distant coast slightly softened, and high-end lifestyle photography composition. Keep the mood quiet and contemplative, centered on reading by the sea; no visible logos, no extra people, no modern clutter, no text overlays.

Create a realistic lifestyle advertising photo in a bright, airy café or sunroom. A young Japanese woman with a short dark brown bob haircut and round eyeglasses sits at a white table, wearing a soft pale blue knit sweater. She rests her chin gently on both folded hands and smiles softly while looking down at a breakfast plate in front of her, conveying a calm “fresh morning” mood. On the table, show exactly 5 main tabletop elements: a white plate with a glossy golden croissant and green salad with grains, a clear glass of lemonade with one visible lemon slice, a small clear vase with leafy stems, a dark brown rectangular placemat, and a fork-and-knife set on the left. The background is softly blurred with warm natural sunlight, large windows or balcony openings, white walls, potted greenery outside, a white dining chair, and one framed ocean-like print on the wall. Use shallow depth of field, high-key natural lighting, gentle lens flare, soft shadows, realistic skin texture, commercial photography polish, and a vertical 2:3 composition. Add small elegant thin serif text near the upper center reading {argument name="headline text" default="目覚めたワタシ、New Born."}. Keep the typography subtle, gray, widely spaced, and integrated like a Japanese wellness or breakfast brand advertisement. Use {argument name="sweater color" default="pale blue"}, {argument name="drink" default="lemonade with a lemon slice"}, and {argument name="breakfast pastry" default="croissant"}.

Using REFERENCE_0 as the base image, keep the castle artwork unchanged and overlay rough handwritten edit notes directly on top of it, as if marking areas for an image-editing revision.

Add exactly 2 annotation groups:
1. Red annotation group on the tall rear tower near the upper-right center: draw red diagonal hatching across the tower, add a red arrow pointing toward it, and add the handwritten Japanese note {argument name="red edit note" default="一番奥の高いこの塔を崩して、無くして空にして。"} in the upper-right sky area.
2. Green annotation group around the crumbling edge/remaining structure near the middle top: draw a green irregular outline around the broken area, add a green arrow pointing down to it, and add the handwritten Japanese note {argument name="green edit note" default="この辺崩れのこっている"} in the upper-middle sky area.

Style constraints: the annotations should look like quick digital pen markup, not integrated into the painting; keep the original watercolor fantasy castle image otherwise intact with no actual architectural removal yet.

Using the provided reference image as the character base, convert the rough sketch into a fully rendered, clean anime-style character illustration while preserving the same head angle, bust portrait framing, facial direction, and general silhouette. Keep the white background and the handwritten name/signature placement on the left.

Transformation: Replace the loose pencil lines with polished line art, smooth cel-shading, glossy highlights, and high-detail coloring. Give the character {argument name="hair color" default="deep red"} hair styled into the same swept-back bun shape, with loose front strands, and make the eyes {argument name="eye color" default="bright green"}. Refine the expression into a calm, elegant, slightly aloof look.

Costume and added details: Add an ornate aristocratic fantasy outfit with a tall cream-and-gold embroidered collar, lace ruffles around the neck, and rich metallic trim. Add exactly 3 prominent jewelry/accessory elements: 1 large gold-and-red jeweled hair ornament on the bun, 1 dangling gold earring with a red teardrop gem, and 1 oval red gemstone brooch at the front of the collar.

Text: Preserve the handwritten sketchbook feel of the left-side annotations, but make them cleaner and dark red. Include exactly 2 visible text blocks: {argument name="character name" default="Mariel Trust"} near the upper left, and {argument name="artist signature" default="IceEmperor"} near the lower left.

Style constraints: High-resolution polished anime illustration, crisp facial features, luminous eyes, detailed hair shine, refined fantasy noble design, no extra background objects, no additional text, no watermark.

Create a luminous, highly detailed semi-realistic anime portrait of one young woman standing at an open rustic wooden window in a Mediterranean hillside village. She has {argument name="hair color" default="soft light brown"} long wavy hair with airy bangs, one dark teal ribbon bow and one small gold butterfly hair clip on the right side of her head, large hazel-brown eyes, delicate porcelain skin, and a calm slightly wistful expression as she turns back toward the viewer. She wears a pale mint-green chiffon dress with a fitted bodice, sheer translucent puff sleeves, a flowing semi-transparent layered skirt, and a thin gold chain belt tied at the waist; keep the fabric elegant, romantic, and softly translucent without explicit nudity. Her pose: standing in three-quarter view, one hand resting on the weathered wooden sill and the other arm relaxed near the open window frame. The scene includes exactly one clear glass vase on the left side of the sill containing five orange daisy-like flowers with green stems and leaves. Frame the composition with two cream lace curtains, one on each side, and an open wooden casement window on the right. Outside the window, show terracotta rooftops, beige stone houses, one church bell tower near the center background, a blue lake or sea inlet, and hazy green mountains under bright morning sunlight. Use soft backlighting, warm golden highlights, gentle depth of field, intricate lace texture, realistic wood grain, delicate hair strands, pastel colors, cinematic natural light, vertical portrait composition, dreamy European summer atmosphere, no text, no watermark.

Goal: Create a vibrant vertical anime key visual / promotional poster for a futuristic virtual idol named {argument name="character name" default="Amaha Mirai"}, blending cute idol energy, sci-fi city branding, and dreamy pastel cyber aesthetics.

Canvas: Vertical 2:3 poster, full bleed, high resolution. Bright daytime sky with saturated cyan blue, fluffy clouds, sparkles, lens glow, floating feathers, flower petals, and soft bloom. Overall palette is white, sky blue, cyan, lavender, pastel pink, and iridescent holographic accents.

Main subject: One central anime girl idol in a dynamic close-up perspective, reaching her left hand toward the viewer with exaggerated foreshortening. She is smiling widely with an inviting open-mouth expression, rosy cheeks, large sparkling blue-purple starry eyes, and long flowing {argument name="hair color" default="silver-white hair with cyan and pink streaks"}. Her hair streams across the image with many ribbons and strands, decorated with futuristic clips including a pink triangular hair clip, black X hairpins, a small white wing ornament, translucent blue-black ribbons, and pastel tech accessories. She wears a white futuristic idol crop top and jacket with translucent sleeves, black straps, holographic trims, wing logo, choker collar, belts, dangling crystal charms, and pink-blue iridescent panels. Make her pose joyful, open, and welcoming, as if inviting the viewer into the future.

Background: A luminous futuristic city plaza with tall glass towers, curved sky rails, floating platforms, large holographic billboards, neon signage, and soft architectural curves. Include one large cylindrical billboard on the left, one rectangular character billboard on the right, and multiple smaller glowing screens integrated into buildings. Add pink flowers blurred in the foreground at the bottom edge for depth.

Typography and text content: Use large Japanese title typography across the top reading {argument name="main Japanese title" default="天羽ミライ"}, with huge elegant white-to-pink gradient characters and small sparkle accents. Directly below it place spaced roman letters reading “AMAHA  MIRAI” with a small feather icon between the words. Add a large handwritten slogan across the lower middle reading {argument name="main slogan" default="きらめく未来に、ログイン。"} in pink and cyan brush lettering on a white glowing swoosh, with small stars and a wing doodle. At the bottom center add “— 近日始動 —”.

Count and placement of visible secondary text blocks: Include exactly 8 secondary text blocks: 1) on the left cylindrical billboard, “VIRTUAL CONNECTS REAL”; 2) left midair Japanese copy “つながる みらい。 一緒に、見に行こう。”; 3) upper right small Japanese copy “夢は、もっと近くにある。”; 4) right rectangular billboard with character portrait, “AMAHA MIRAI” and “More Kawaii More Future” plus a heart; 5) right mid-lower Japanese copy “バーチャルから はじまる、 本物のキセキ”; 6) lower left vertical icon list with four items: “MUSIC”, “GAME”, “TALK”, “LIVE”, followed by “— AND MORE...” ; 7) tiny lower-left project label “AMAHA MIRAI Vtuber Project” with small social-style icons; 8) tiny lower-right tagline “NEW STORY. NEW WORLD. WITH YOU.”

Visual style: Polished high-detail anime illustration, idol poster composition, glossy gpt-image style, crisp linework, translucent fabric, holographic reflections, sparkling eyes, soft atmospheric depth, cinematic bloom, clean commercial design, cute and futuristic.

Constraints: Keep exactly one main character. Keep the poster crowded but readable. Preserve all specified text placements and approximate wording. Avoid realism, dark mood, gritty textures, extra characters, distorted hands, unreadable main title, or random additional logos.

Create a cinematic photorealistic full-body cosplay portrait of {argument name="character name" default="Soara"}, a young fantasy craftswoman adventurer standing confidently in a low-angle hero pose on a grassy rocky hillside. She has long {argument name="hair color" default="silver-gray"} hair tied into a high side ponytail with a large dark brown bow, soft bangs framing her face, pale skin, and a calm serious expression. Dress her in a dark forest-green military-style tunic dress with gold piping, high collar, long sleeves with wide cuffs, double rows of brass buttons, a brown leather belt, shoulder straps, black textured tights, and tall dark brown leather boots with light shearling cuffs. Add exactly 5 visible carried/tool elements: 1 large adjustable wrench in her left hand, 1 wooden mallet in her right hand, 1 blue leather tool pouch at her hip holding metal hand tools, 1 brown backpack strapped to her back, and 1 cluster of metal tools protruding from the backpack. The setting is a whimsical medieval-fantasy artisan village beside a massive ancient tree, with a cozy workshop doorway carved into the tree trunk on the right, a small round-roofed cottage and wooden balcony structures in the midground, scattered barrels and workshop clutter, distant green mountains, rocks, grass, and a bright cloudy sky. Use natural warm daylight, soft haze, shallow depth of field, highly detailed fabric and leather textures, realistic skin, wind-tossed hair strands, muted earthy colors, epic fantasy realism, vertical 4:5 composition, no text, no watermark, no anime line art.

Create a full-body watercolor anime character illustration on a tall vertical canvas with an off-white paper background, thin black outer border, and scattered gray and orange paint splatters. The subject is {argument name="character name" default="ミシェル"}, a small fantasy Halloween kemonomimi girl standing front-facing with a slight sleepy open-mouth expression, golden-orange eyes, rosy cheeks, and long messy {argument name="hair color" default="pink"} hair. She has exactly 2 large fluffy fox/cat ears on top of her head and 1 thick pink scaly dragon-like tail curling behind her legs. Dress her in an oversized tattered {argument name="coat color" default="orange"} Halloween jacket with black patchwork, ragged hems, zipper details, straps, buckles, small pumpkin-face patches, and a dangling pouch; underneath, show a black layered dress or skirt with gothic ruffles and a dark high-collar top with a small pendant. Add an asymmetrical pointed hood or jester-witch hat behind one ear in orange and black, decorated with a round eye-like emblem, a skull motif, and exactly 2 dangling charms: one small skull and one triangle. Add small bat wings attached behind the coat, exactly 2 wings total, orange with black edges. Footwear: exactly 2 chunky black lace-up boots with orange trim, mismatched socks, one orange-striped sock and one black fishnet sock. In the background, include exactly 3 separate flying bats: one dark bat on the upper right, one pale gray bat on the left, and one tiny dark gray bat near the upper left, plus a few star-shaped ink marks. Use delicate ink linework, soft watercolor washes, transparent layering, textured paper grain, and a cute dark-fantasy Halloween color palette of orange, black, gray, cream, and pink. Place a small handwritten signature reading {argument name="signature text" default="ミシェル"} at the bottom right. Avoid photorealism, avoid extra characters, avoid additional readable text, and keep the composition centered like a character concept art poster.

Create a photorealistic cinematic news-style image of a massive humanoid police robot from the Tokyo Metropolitan Police heavy machinery response unit standing on a wet elevated expressway in Tokyo, carefully carrying one black-and-white Holstein cow in both armored hands as if rescuing or detaining it. The robot is about three stories tall, built from white and dark navy modular armor plates with visible bolts, hydraulic joints, scuffed paint, rain grime, red emergency light bars on both shoulders, thin antennae on its helmet, and Japanese police markings reading {argument name="police marking text" default="警視庁"} on the chest and shoulder armor. The cow should be full-sized, dangling but unharmed, with four visible legs, black patches, a white face stripe, yellow ear tag, and a confused expression. Behind the robot on the left, include exactly one futuristic armored police vehicle with six large wheels, black-and-white livery, roof equipment, small red-and-blue lights, headlights on, and Japanese police markings. Set the scene on a rainy urban highway with reflective asphalt, lane markings, guardrails, and a green overhead road sign showing Shibuya and Tanimachi with an exit distance, plus blurred high-rise buildings in the background. Use an overcast stormy gray sky, misty atmosphere, damp surfaces, realistic scale, low-angle street-level perspective, 35mm documentary photography look, sharp foreground detail, subtle motion realism, dramatic but plausible lighting, no cartoon style, no extra animals, no extra robots, no people, no watermark.

Create an epic dark fantasy cinematic battle scene of {argument name="main creature" default="a colossal minotaur"} assaulting {argument name="fortress type" default="a medieval stone fortress"}. Use a dramatic low-angle perspective from the viewpoint of human soldiers at the creature’s feet, making the monster tower over the frame. The minotaur is extremely muscular, covered in rough brown hide and tangled black fur around the neck and head, with two huge curved ivory horns, a snarling bull face, bared teeth, and furious eyes. It wears brutal dark leather and iron barbarian armor: wrist bracers, straps across the chest, metal rings, a ragged leather loincloth, belts, and spiked details. It raises both arms overhead, gripping a massive double-bladed battle axe with a long wrapped handle and a weathered, chipped, dark steel blade, preparing to smash downward. In the foreground, show exactly two human defenders partially visible from behind: one armored soldier at lower left with a rounded steel helmet, shoulder armor, sword, and red cloak, and one armored soldier at lower right with a steel helmet and dark armor. Around them, depict flying chunks of stone, dust, gravel, and impact debris suspended in the air. The background shows tall ruined castle towers, battlements, arched stone walls, tattered red banners, smoke, haze, and bright cloudy daylight breaking through, creating harsh rim light and deep shadows. Visual style: ultra-detailed photorealistic fantasy concept art, gritty medieval realism, cinematic composition, dynamic action, high contrast, textured stone and metal, dust-filled atmosphere, shallow depth of field, dramatic scale, no modern objects, no text, no watermark.