Abstract / Background

Extreme close-up, highly detailed Blender + Octane render of the attached product images arranged as a dense, overlapping 3D floating cluster. Use the exact products shown in the attached photos preserve their true colors, materials, logos, labels, textures, and proportions exactly as photographed. Do not invent, substitute, or generically reinterpret any product; render them as faithful 3D versions of the attached references. Composition: objects overlap and float in a tight macro cluster, each catching light from a different angle, creating a layered depth-of-field effect with the nearest object in sharp focus and others softly falling off in the background blur. Intertwined throughout the cluster: long, curling streamers of metallic confetti ribbon and thin silky ribbon, threading between and behind the objects, catching specular highlights as they curl through the frame. Materials and rendering: emphasize contrasting textures true to each attached product brushed or polished metal, soft fabric sheen and folds, glass with complex caustics and refractions if transparent liquid is present, waxy or matte surfaces, glossy plastic or lacquer all rendered with physically accurate reflections, refractions, and soft shadow occlusion. Background: a sophisticated multi-tonal gradient, built creatively from the dominant and accent colors actually present in the attached product images analyze the attached photos' palette (primary object color, packaging color, any metallic or accent tones) and construct a smooth, moody gradient that transitions between 2–3 of those extracted tones, deepest/coolest tone at the bottom edge, warmest/lightest tone at the top edge, evoking a refined evening luxury atmosphere. Do not default to a fixed palette the gradient must visibly relate to and complement the attached products' actual colors. Confetti ribbon colors should also pull from this same extracted palette (one metallic tone + one soft tone). Lighting: soft key light from the upper background glow, subtle rim lighting on each object edge, gentle bounce light between overlapping items, no harsh shadows. Camera: macro lens look, shallow depth of field, extreme close-up framing, photorealistic, ultra-detailed, 8K render quality.

Convert a user's short image request, attached image, or combination of text and images into one finished prompt for GPT Image Gen V2 or Nano Banana Pro. Apply the Redphase Survey visual system defined below. Do not explain the system, show analysis, or add commentary. Return only the finished image prompt in the selected model's format. Use the model named by the user or supplied by the host application. If neither names one, use GPT Image Gen V2. For GPT Image Gen V2, return natural language prose only. For Nano Banana Pro, return one valid JSON object only. When both are explicitly requested, return the GPT prompt first under the line "GPT IMAGE GEN V2", then the JSON under the line "NANO BANANA PRO". Add no other text. Write prompt content in the user's language unless another language is requested. Keep Nano Banana Pro keys in English. Preserve exact in-image text in its required script. CORE LOGIC Every image contains two simultaneous records. The survey record describes what occupies the scene. It uses observed geometry, dry line, hatching, reserved paper, and a restricted set of local colors. The red phase describes what is active through the scene. It uses one red spectral range, deep black, limited bloom, rapid falloff, and selective loss of ordinary color and detail. The survey record answers: What is present? The red phase answers: What is acting, passing, gathering, or changing? Do not average the two records into red-tinted illustration. Give each record a separate job. Before writing the final prompt, silently determine: 1. The literal subject and setting. 2. The stable material structure that belongs to the survey record. 3. One active condition that belongs to the red phase. 4. The source of that condition. 5. The path or relation through which it travels. 6. The receiving subject, object, region, or surface. 7. The visible consequence. 8. Two or three transition sites where the records meet. Choose the active condition from the user's request. It may be physical, social, temporal, or perceptual. Valid causes include heat, movement, wind, electric current, pressure, recent touch, reflected light, sound, gaze, attention, occupancy, growth, chemical change, distance, or another condition that can be shown through visible evidence. If the request names no active condition, use the strongest existing process already implied by the subject or action. Do not invent danger, desire, trauma, memory, menace, or supernatural activity without evidence. Red is a measurement channel, not a mood label. Do not equate red with violence, horror, romance, or urgency unless the user's content establishes that meaning. The active condition must have one source, one route or relation, and one consequence. Do not scatter red accents without a causal structure. CONTENT CONTROL Preserve the user's explicit subject, count, identity requirements, action or state, setting, era, clothing, objects,

{"vibe_title_en": "Ethereal Surrender in the Digital Baroque", "master_prompt": "The Deconstructed Detail framing, 100mm Macro f/1.2 lens, Hasselblad H6D-100c. A hyper-realistic, photographic surrealism shot focusing exclusively on the cheekbone and jawline of The Protagonist, treating the body as a canvas. The subject's head is tilted back in 'The Ethereal Surrender', radiating a peaceful, dreamlike submission. A surreal, high-budget practical effect rests across the upper face: an ornate, baroque-style blindfold woven entirely from glowing fiber-optic filaments, emitting a soft, holographic powder-blue light that physically illuminates the peach fuzz and microscopic pores of the skin. A single, perfect drop of liquid gold is caught mid-roll down the cheek. Editorial styling: flawless, dewy glass-skin makeup with a flushed peach lip stain. Extremely long, heavy jet-black hair is styled in a severe, glossy wet-look, slicked cleanly behind the ear, cascading out of frame over a bare shoulder. The background is completely dissolved into a creamy, high-end f/1.2 bokeh of a grand, decaying 18th-century baroque hall lit by hundreds of melting beeswax candles, creating rich, golden amber orbs of out-of-focus light. Lighting is a masterful cinematic collision: the warm, directional Rembrandt glow from the background candles against the icy, practical blue illumination of the optical lace pressing into the cheek. Authentic Kodak Portra 400 film stock texture, tangible physical weight, absolute optical fidelity, no plastic CGI feel. NO NEON.", "meta": {"intent": "Macro surrealism combining classical baroque opulence with futuristic digital textures", "priorities": "Macro detail, extremely shallow f/1.2 bokeh, hyper-realistic skin texture, practical light contrast", "device_profile": "Hasselblad H6D-100c, 100mm Macro Prime Lens"}, "frame": {"aspect": "4:5", "composition": "The Deconstructed Detail, focusing entirely on the cheekbone, jawline, and the edge of the surreal blindfold, ignoring the full face", "layout": "Off-center macro focus, dominant negative space filled with golden candlelight bokeh", "camera_angle": "Extreme close-up profile, slightly tilted up from below to emphasize the 'Ethereal Surrender' posture", "tilt_roll_degrees": "15 degrees"}, "subject": {"gender": "Female", "identity": "The Protagonist", "demographics": "Androgynous, ageless editorial model", "face": "Hyper-detailed skin pores, peach fuzz, glossy glass-skin makeup finish, flushed peach lip tint, a single rolling tear of heavy liquid gold", "hair": "Extremely long, heavy jet-black hair, styled in a severe, glossy wet-look, slicked perfectly behind the ear and cascading down", "body": "Bare shoulder visible at the edge of the frame, relaxed neck muscles showing a state of peaceful submission", "expression": "Serene, weightless, lips softly parted in sensory, dreamlike surrender", "pose": "Head tilted back, eyes closed (hidden under the accessory), neck exposed"}, "wardrobe_accessories": {"garments": [{"item": "Bare skin", "material": "Human skin treated as a canvas", "color": "Natural skin tone", "fit": "N/A"}], "accessories": [{"item": "Baroque blindfold visor", "color": "Holographic powder-blue", "material": "Woven fiber-optic filaments behaving like 18th-century lace", "brand_style": "Avant-Garde High-Tech Historical"}, {"item": "Tear", "color": "Metallic gold", "material": "Liquid gold", "brand_style": "Surreal practical effect"}]}, "environment": {"setting": "A decaying, opulent 18th-century baroque opera box (entirely out of focus)", "surfaces": "Hyper-textured human skin, glowing optical threads, soft out-of-focus background architecture", "depth": "Extremely shallow depth of field (f/1.2), isolating only the cheek and the optical lace", "atmosphere": "Warm, floating, dreamlike, thick with the tangible heat of hundreds of unseen candles", "lens_interaction": "Creamy, spherical bokeh from background flames, soft blooming glow from the fiber optics"}, "lighting": {"key": "Practical holographic powder-blue glow emitting physically from the fiber-optic blindfold", "fill": "Warm amber candlelight reflecting from the out-of-focus background", "rim": "Soft golden highlight tracing the jawline and the glossy wet hair", "shadows": "Deep, rich, and velvety warm shadows with high micro-contrast at the focus plane", "color_temperature": "Mixed: 2700K (Candles) contrasting with 6500K (Fiber-optic blue)", "sensor_flare": "Subtle halation around the glowing blue filaments resting on the skin"}, "camera": {"lens_type": "Macro Prime", "focal_length": "100mm", "aperture": "f/1.2", "focus": "Pin-sharp exactly on the intersection of the glowing blindfold, the skin pores, and the liquid gold tear", "sensor_format": "Medium Format Digital", "perspective_distortion": "Zero distortion, highly compressed background elements"}, "post_processing": {"color": "Rich analog color grading, warm golden shadows, striking pastel blue highlights", "tonality": "High micro-contrast on the focal point, extremely soft fall-off into the bokeh", "texture": "Authentic Kodak Portra 400 grain, rendering microscopic details with organic warmth", "digital_sharpening": "None, relies entirely on elite optical sharpness", "chromatic_aberration": "Very subtle, natural fringe at the very edge of the macro depth of field"}, "negative_specifications": ["neon", "full body shot", "wide angle", "CGI feel", "plastic skin", "harsh studio flash", "modern urban settings", "short hair", "matte makeup", "digital illustration", "flat lighting", "eyes open"]}

Create a cinematic 16:9 matte deep-blue sci-fi wallpaper of a vast futuristic megacity at night, seen from a distant elevated viewpoint across dark water or haze. The right half of the frame is dominated by immense vertical skyscraper silhouettes, needle towers, and dense black architecture, with the tallest structures fading into mist; the left half is mostly open stormy blue-black sky for negative space. Treat light like paint: orange ember-like windows and cyan-white glows are stippled, dragged downward into vertical streaks, woven textures, and reflective distortions across the tower faces. Add heavy atmospheric fog, subtle film grain, slight defocus, soft bloom, rain-slick reflections, and long vertical smear artifacts, giving the city a matte, moody, painterly, dystopian aesthetic. Include exactly 6 flying light craft or aircraft trails: 1 prominent sleek craft near the upper center moving diagonally toward the right with two long pale contrails, and 5 much smaller distant streaks scattered low across the left and middle horizon. Use a restrained palette of navy, midnight blue, black, muted violet, cold cyan, and sparse burnt orange highlights. No people, no readable text, no logos, no borders, no sharp clean CGI look; make it feel like a grainy concept-art frame from an atmospheric science-fiction film.

Create a dramatic wide cinematic digital collage inspired by Baroque opera and classical mythology: in the center, a terrified young mother in a torn cream linen dress rises from a violent dark teal ocean while clutching her small curly-haired child protectively against her chest, both lit with painterly golden highlights and looking upward in fear and defiance. Behind them, looming in the upper left, place a monumental stone-gray goddess bust representing Juno, wearing an ornate crown and radiant fan-like halo, with one huge raised hand emerging from storm clouds as if controlling the tempest. Surround the scene with crashing waves, sea foam, wind streaks, splashes of gold paint, and swirling engraved music-staff lines, creating the feeling that the ocean, wind, gods, and music are all fused together. Include exactly five visible musical elements: two long Baroque flutes crossing diagonally from the left edge, two golden natural horns floating in the upper right, and one abstract vertical lyre/string motif along the right edge. Add decorative red and pale pink treble-clef-like curls and circular staff spirals woven into the waves. Use a dark palette of deep navy, teal, black, antique gold, ivory, coral red, and weathered stone gray. Style: ultra-detailed cinematic mythological poster, mixed-media oil painting plus etched musical notation, turbulent water realism, dramatic chiaroscuro, high contrast, epic album-cover composition, no readable text, no modern clothing, no watermark. Aspect ratio 16:9.

4:5 A photorealistic luxury editorial portrait seamlessly integrated into a flowing liquid-chrome ribbon with shifting iridescent reflections of sapphire blue, emerald green, violet, and molten gold. The ribbon twists into impossible Möbius-like loops while the portrait remains perfectly upright from every visible angle, creating an impossible visual paradox. Natural skin tones with premium cinematic color grading. Deep charcoal background. Minimal composition. Luxury fashion campaign. Soft volumetric studio lighting. Mirror-like reflections. Ultra realistic. 85mm lens. Editorial masterpiece. Hyper-detailed. No text, no watermark.

Goal: Create a polished iOS-style wellness app mockup for Lumen, a light-themed mood and breathing tracker, with soft Apple-inspired glassmorphism and Japanese UI text. Canvas: Vertical mobile screenshot, 9:16 aspect ratio, approximately 563 × 1008 px. Use an airy cream-white background with a misty watercolor landscape at the bottom and edges, subtle mountains, reeds, and a glowing winding light trail running vertically through the center in warm peach near the top and pale blue near the bottom. Layout: Place the app title at top left in a refined serif font, large black text reading “Lumen”, with the subtitle “今日の光” underneath in smaller dark gray Japanese text. Place one circular hamburger menu button at top right with three dark horizontal lines and a soft shadow. Main UI elements: Use exactly 5 floating rounded glass cards/buttons stacked vertically down the center: 1. Top small pill card: warm peach sun icon on the left, label “集中”, time “10:20”, and one small orange status dot on the right. 2. Middle selected pill card: larger white translucent pill with teal glow outline, teal circular wave/breath icon on the left, label “穏やか”, time “15:40”, and one small teal status dot on the right. Add a small downward notch centered at the bottom of this pill, visually connecting to the expanded card below. 3. Expanded mood entry card: large rounded white translucent panel. Top left has the same teal wave icon, title “穏やか”, and a circular close button with an “×” at top right. Show the time “15:40” in large text. Include one horizontal mood slider with exactly 3 visible parts: a sad face icon on the left, a teal-to-gray slider track with the knob slightly right of center, and a happy face icon on the right. Beneath it, include a rounded text input containing “深呼吸のあと、少し軽くなった” with a small pencil icon on the right. At the bottom, include a large teal gradient rounded button labeled “記録する”. 4. Lower small pill card: blue moon-and-star icon on the left, label “回復”, time “20:45”, and one small blue status dot on the right. 5. Bottom navigation / action bar: wide translucent rounded capsule near the bottom. It contains exactly 3 circular controls: left small circular button with a bar chart icon, center larger glowing golden circular button, and right small circular button with a heart outline. Under the center glowing button, show the label “気分を追加”. Visual style: Minimal premium iOS app design, translucent frosted glass surfaces, rounded corners, soft drop shadows, delicate glow, subtle depth, clean spacing, calm wellness atmosphere. Use warm peach, mint teal, pale blue, cream white, and charcoal typography. Icons should be simple line icons and feel native to iOS. Constraints: Keep all visible Japanese text exactly as specified. Do not add extra cards, extra text, logos, watermarks, device frame, or status bar. The mockup should look like a generated app UI concept rather than a screenshot of a real installed app.

Goal: Create a vertical pixel-art sci-fi music sequencer interface showing four cute monster audio tracks, like a retro game synth dashboard. Canvas: Tall portrait mobile screen, 720×1200 style, dark navy background with glowing rounded rectangular panels and crisp pixel-art edges. Use a limited neon palette with cyan, mint, red, purple, and yellow accents. Everything should look like high-resolution pixel art, with subtle scanline/dither texture and no photorealism. Layout: At the very top, create one wide rounded header bar. Inside it place exactly 6 discrete control/readout elements: 1 small alien/robot face icon on the left, 1 dotted equalizer readout beside it, 3 square transport buttons in the center showing stop, play, and record icons, and 1 dotted menu/grid indicator on the right. Main track area: Stack exactly 4 horizontal rounded track panels, each with a left track number and a vertical level meter, a monster character on the left, a circular dotted status control in the middle, and a large rotary knob on the right. Track 1: Label number “1”. Mint/teal theme. The creature is a sleepy squat round monster with small horns/spikes, closed eyes, lying down, with “Z Z” sleep marks above it. The middle control is a dotted circular ring with a small center dot. The right knob has a pale yellow arc and one mint indicator dot near the top. Track 2: Label number “2”. Red theme. The creature is a red one-eyed monster standing aggressively, mouth open with teeth, one arm raised and casting a crackling magical red orb above its hand. The middle control is a red dotted ring with a solid red center dot. The right knob has a red arc and one red indicator dot. Track 3: Label number “3”. Cyan theme. The creature is a blue dancing one-eyed monster wearing headphones, smiling with a toothy mouth, arms raised, with exactly 2 floating music notes near its head. The middle control is a cyan dotted circle containing a play triangle. The right knob has a cyan arc and one cyan indicator dot. Track 4: Label number “4”. Purple theme. The creature is a purple energetic monster with multiple ghosted afterimage duplicates behind it, big eyes, waving arms, small sparkle marks around it. The middle control is a purple dotted circle containing a play triangle. The right knob has a purple arc and one purple indicator dot plus a tiny triangular tick at the top. Bottom mixer section: Create one wide rounded bottom panel. On the left, include exactly 1 square album-art tile with a glowing mint border, showing a floating mossy cube planet or asteroid with small orbiting rocks and tiny stars. To the right of the album art, arrange exactly 5 square effect buttons in one row: 1 purple concentric-circle/radar icon, 1 cyan vertical waveform/spectrum icon, 1 green pixel burst icon, 1 yellow audio waveform icon, and 1 yellow ringed planet icon. Beneath those buttons, add exactly 2 horizontal audio meters: the left meter has a microphone icon followed by green-to-yellow vertical bars, and the right meter has a speaker icon followed by green-to-yellow vertical bars, separated by a thin vertical divider. Text content: Only include the visible track numbers “1”, “2”, “3”, and “4”; do not add readable labels or words besides the two “Z” sleep marks. Use abstract pixel icons rather than detailed typography. Style constraints: Retro 16-bit pixel-art UI, dark futuristic synth console, glowing outlines, rounded panels, clean symmetry, high contrast, playful monster characters, no watermark, no realistic 3D rendering. Make the main accent color palette customizable: mint cyan, hot red, purple, deep navy, retro monster music sequencer.

Goal: Create a realistic vertical smartphone screenshot of a ChatGPT-style conversation showing a user asking for a promotional bonus banner and the AI returning the finished banner image. Canvas: 9:16 mobile screenshot, approximately 586 × 1024 px, white background, iPhone-like status bar at the top showing 7:17, 5G, and a low battery indicator. Minimal clean app UI with soft shadows and rounded controls. Layout: At the top left place a circular hamburger menu button. At the top right place a floating rounded pill with two icons: an edit/pencil icon and a three-dot menu. In the upper middle-right, show a cute anime-style boy avatar with curly dark brown hair, large sparkling eyes, rosy cheeks, and a black hoodie, resting his cheek on one hand. Below him place one large rounded light-gray chat bubble with Japanese prompt text. Beneath the bubble place one wide horizontal promotional banner image, scaled to fit the phone width. Under the banner add small app action icons, then a bottom input bar with a plus icon, placeholder text, microphone icon, and green voice button. Chat bubble text: Use Japanese text exactly as visible, with the first phrase in blue: 「画像を作成する」 followed by black text: 「公式LINE登録で今だけもらえる7大特典というバナーを横長で作成して。ジャンルは対面営業完全攻略。添付の写真はコンテンツホルダーなので必ず使用して。JSON形式でプロンプトを出力してから画像生成して」 Banner design: Create a high-conversion Japanese lead-magnet banner for official LINE registration. Use a luxury black, navy, and gold color scheme with bright highlights, diagonal metallic accents, and a premium sales-promotion feel. The banner headline should read 「今だけもらえる 7大特典」 with a very large gold “7” and bold white/gold typography. Include the topic subtitle 対面営業完全攻略. Include the same anime boy avatar on the right side of the banner. Banner sections: Show exactly 7 benefit cards in a row across the bottom, each with a small thumbnail image, a numbered label, and Japanese title text. The seven visible cards are: 1)営業の勝負所を完全攻略, 2)アプローチトーク, 3)提案力強化テンプレート, 4)クロージング実践トーク集, 5)営業力UPチェックリスト, 6)営業ロールプレイ実演動画, 7)トップ営業のセット. Add a green LINE-themed call-to-action strip along the bottom reading 「今すぐLINE登録して豪華 7大特典を受け取る!」 with a small green LINE icon. Visual style: The phone UI should look like a real screenshot, while the avatar and banner should be polished AI-generated digital art. Keep the chat interface spacious and uncluttered, with realistic rounded corners, subtle shadows, and crisp Japanese typography. Constraints: Use exactly one chat bubble, exactly one avatar above the bubble, exactly one banner image, exactly seven benefit cards in the banner, exactly two icons in the top-right floating pill, and exactly four controls in the bottom input bar. Do not add extra messages, watermarks, or unrelated text.

A bold geometric cubist portrait illustration with fragmented polygonal color planes and faceted angular shapes, inspired by modern pop-art and abstract expressionism. Dynamic digital painting featuring crisp sharp-edged brushstrokes mixed with subtle textured paint overlays, creating a handcrafted acrylic-on-canvas appearance. Strong cinematic high-contrast lighting with dramatic rim light and deep shadows. A vibrant complementary color palette dominated by deep navy blue, crimson red, fiery orange, golden yellow, bright cyan, and cool teal, balanced with warm skin-toned highlights. Clean edge separation, bold color blocking, expressive facial planes, and stylized geometric shading without heavy outlines. Rich digital paint texture with visible brush grain, layered pigment effects, and slightly rough canvas surface. High color saturation, exceptional tonal contrast, ultra-detailed rendering, sharp focus, futuristic editorial aesthetics, modern gallery-quality illustration, premium poster design, cinematic atmosphere, dark minimalist background for maximum visual impact, 8K ultra-high resolution.

Goal: Create a high-impact Japanese YouTube thumbnail about automatically generating video captions/telops with AI, featuring a dramatic comparison from Vrew to Claude Code. Canvas: 16:9 horizontal thumbnail, 1200×675 style, optimized for YouTube. Use a dark blurred video-editing timeline interface as the full background, with purple and blue timeline tracks and small clip blocks visible but out of focus. Layout: Put a huge bold Japanese headline across the top reading AIでテロップを自動作成. The letters should be white with thick black outline and an additional thin white outer stroke, slightly tilted and tightly spaced for maximum impact. Place one large white speech-bubble complaint box on the left-middle with a black outline and a small pointer at the bottom right. Inside it, use bold black Japanese text for the complaint and a red emphasized final line with a yellow underline: 誤字脱字多すぎるし、テロップはズレまくってるし、変なところで改行するし、そのくせにそれ以外のことは何もできないって...\nVrewマジで使えんw. Place two app-style logo tiles in the center-right: exactly 2 logos total. Logo 1 is a cyan-blue rounded square Vrew-like icon with a white V mark, located near the center, partially overlapped by a small orange pixelated mascot/bug shape. Logo 2 is a larger orange-red rounded square Claude-like icon with a white radial starburst, located to the upper right, with a shiny gold crown tilted on its top-right corner. Add a small white label under the orange icon reading Claude Code in bold black text. Between the two logos, add exactly 1 curved red arrow with a white outline pointing from the blue logo toward the orange logo. Bottom text: Fill the bottom third with one enormous Japanese conclusion phrase reading これで全て解決. Make it thick rounded display lettering with a bright yellow gradient fill, heavy black outline, and glowing white outer stroke. The text should overlap the background and partly sit in front of the center graphics, creating a loud thumbnail composition. Visual style: Japanese viral tech thumbnail, extremely bold contrast, dense composition, exaggerated outlines, drop shadows, white glow effects, saturated colors, black background atmosphere, urgent and humorous tone. Make all Japanese text crisp, legible, and correctly spelled. Discrete elements to include: exactly 1 top headline, exactly 1 speech bubble, exactly 2 app logos, exactly 1 orange mascot/bug shape, exactly 1 red curved arrow, exactly 1 gold crown, exactly 1 Claude Code label, and exactly 1 huge bottom headline. Do not add extra logos, people, watermarks, or unrelated text.

Creative 3D ad for [BRAND], with surreal object made from it, matching background color, real slogan below, logo on top, miniature person interacting, minimal and clever concept

Treat the uploaded image as composition and color reference only. Recreate it as an original minimalist vector doodle illustration. Use thick monoline cobalt-blue outlines with intentionally uneven hand-drawn strokes. Fill every object with completely flat solid colors sampled from the reference. No gradients, no realistic shading, no marker texture, no pencil texture, no cross-hatching, no painterly effects, no photorealism. Simplify facial features into cartoon shapes with round eyes, tiny nose, simple mouth, and minimal details while keeping the pose instantly recognizable. Use a clean off-white paper background with lots of negative space. The overall style should match a modern editorial doodle illustration, similar to hand-drawn vector artwork created with Procreate or Adobe Fresco, playful, bold, flat, minimal, and graphic Do not create a marker sketch. Do not create a realistic illustration. Create a flat vector doodle with bold blue outlines exactly in the same visual language as the reference.

Goal: Create a square Chinese dark newspaper-style story poster about 代付链接, using a restrained, emotional, old-magazine collage aesthetic. Canvas: 1:1 square poster, black worn paper background with heavy grain, scratched edges, dust, scuffs, faded ink, photocopy texture, and uneven white distressing. Overall monochrome black, gray, and off-white only. Layout: Irregular two-column editorial layout beneath one huge distressed headline. The headline occupies the top third, aligned slightly left, in oversized rough white Chinese brush/newspaper type reading 代付链接. Below it, arrange exactly 11 discrete text blocks: 6 blocks in the left column and 5 blocks in the right column. Use a mix of plain paragraph blocks, thin hand-drawn rectangular outlines, torn-paper strips, and large quoted pull-quotes. Keep margins tight and imperfect, like an underground zine or old crime newspaper clipping. Text blocks, counted exactly: 1. Left upper plain paragraph: “第一次加上她微信，\n是我姐的邻居牵的线。\n\n邻居是女孩的姑姑，说女孩二十五六岁，\n在老家市里卖车险，人长得漂亮，性格也不差，\n平时接触的人少，所以想找个靠谱的对象。” 2. Left outlined box paragraph: “我对“靠谱”这两个字没抱太大希望。\n相亲见得多了，介绍人口中的“性格好”，\n有时候只是“不爱发脾气”；“工作稳定”，\n可能只是暂时没辞职；“长得显小”，\n多半意味着照片拍得不错。” 3. Left bold intro paragraph: “但她的照片确实好看。\n脸小，皮肤白，笑起来显得很年轻，\n乍一看比实际年龄小好几岁。\n我们刚开始聊得也还算顺畅。\n她会问我工作，也会主动发些日常，\n偶尔说一句俏皮话。” 4. Left torn-paper strip: “我那时候觉得，也许这次能正常一点。” 5. Left large outlined pull-quote: ““又是一个代付链接。”” 6. Left lower plain paragraph: “我把钱付了。\n然后告诉自己，也许只是她性格大大咧咧，\n也许只是试探，也许现在年轻人相亲就这样。” 7. Right upper small lead-in plus torn-paper quote: “没过几天，她忽然发来一句：” followed by a torn strip reading ““请我吃个早饭呗。”” 8. Right upper paragraph: “我以为她是在开玩笑，或者想让我\n给她点份外卖，于是刚准备问她想吃什么，\n一个代付链接已经甩了过来。\n干脆、直接，没有铺垫。\n我盯着那个链接看了几秒，\n心里某个地方轻轻往下一沉。” 9. Right middle torn-paper quote: ““我们还没见过面。”” 10. Right middle paragraph with outlined lower portion: “甚至连彼此说话的语气都没完全熟悉，\n她已经很自然地让我替她付款。\n金额不大，几十块钱。\n可让人不舒服的，\n从来不是那几十块，而是那种理所当然。\n\n我没有立刻发作。\n说到底，还是因为她长得漂亮。\n人有时候很诚实，也挺没出息。\n要是换成一个完全不符合自己审美的人，\n可能当场就不想聊了。可面对一个漂亮女孩，\n很多原则都会先被自己往后挪一点。” 11. Right lower closing paragraph plus torn-paper emphasis: “过了一两天，聊到晚上，她忽然又说想吃牛排。\n紧接着，又是一个代付链接。\n这一次，我连替她找理由的力气都没有了。” followed by a gray torn-paper block reading “我没付。\n她也没生气，只是发了个表情，\n好像这件事本来就无关紧要。\n可我心里已经把她划掉了。” Typography and style: Use bold, rough, distressed Chinese fonts with chipped edges. Main headline should look hand-painted and eroded. Body text should be white-gray, clear enough to read, with uneven baseline and analog print texture. Pull-quotes should be much larger than body text, inside rough rectangular outlines or pasted torn-paper frames. Add hand-drawn border lines, faint column rules, scratches, smears, and random paper abrasions. Constraints: Preserve the exact Chinese text as much as possible, avoid adding new text, avoid color, avoid clean modern design, avoid illustrations or people, no logos, no watermark.

Draw with very simplified lines and design, with the quality of a doodle written in a textbook, on a piece of crumpled paper.

A simple blue circle centered on a plain white background, no text, no watermark

Add a large amount of flowers and a large amount of weapons to the background, make it look cool overall, vertical 3:4

Create a breathtaking vertical cinematic fantasy artwork on a dark star-filled cosmic background. Show exactly two figures: a tiny solitary human silhouette standing at the bottom center on a reflective, faintly rippled surface, head bowed, and a colossal translucent feminine celestial spirit rising directly above him like a luminous apparition. The spirit is human-shaped from head to flowing lower body, with a serene upward-facing profile, arms relaxed at her sides, and her form made of glowing nebula filaments, constellations, plasma veins, and sparkling particles. Use electric blue and cyan light for the outer body, intense golden-orange light for the inner energy streams, and deep violet-purple smoke trails flowing backward from her head and upper body toward the upper left. The lower part of the spirit should emerge from a bright explosion of golden sparks and swirling energy at the human’s feet, as if imagination or consciousness is materializing into a cosmic goddess. The composition should be centered and dramatic, with strong scale contrast between the small person and the immense radiant apparition. Style: ultra-detailed digital art, high-contrast astrophotography look, volumetric glow, luminous particle effects, ethereal transparent anatomy, elegant flowing energy ribbons, magical realism, awe-inspiring, sharp focus on the figure, deep black space with scattered stars. Add a subtle handwritten signature reading Kollab in the bottom right. Use electric blue, golden orange, and violet purple. The small human should appear as a bowed young man in dark clothing, and the giant apparition should appear as a serene feminine cosmic spirit. No borders, no UI, no extra text, no logos other than the small signature.