[{"data":1,"prerenderedAt":68},["ShallowReactive",2],{"detail-typography-portrait-art-325":3},{"image":4,"related":18},{"id":5,"url":6,"thumb":6,"prompt":7,"prompt_cn":8,"prompt_en":7,"tags":9,"title":13,"source_repo":14,"date":15,"width":16,"height":16,"slug":17,"urlWebp":6,"thumbWebp":6},325,"\u002Fimage\u002F21c67e14442843a7.webp","A high-energy VTuber thumbnail illustration of a smiling anime girl with {argument name=\"hair color\" default=\"bright blue\"} hair in a high ponytail wearing a white shirt. The background is an explosive burst of rainbow light rays and golden sparkles. A golden retro microphone sits in the bottom left. Massive, shiny 3D gold text on the left reads \"{argument name=\"main title text\" default=\"初配信\"}\". A 3D gold and blue subtitle reads \"{argument name=\"subtitle text\" default=\"一緒に最高の時間を！\"}\". An ornate blue and gold oval badge in the bottom right displays \"{argument name=\"character name\" default=\"エリン Erin\"}\". A red top-right badge reads \"{argument name=\"badge text\" default=\"LIVE\"}\".","",[10,11,12],"Typography","Portrait","Art","Colorful VTuber Debut Stream Thumbnail","https:\u002F\u002Fgithub.com\u002Ffreestylefly\u002Fawesome-gpt-image-2.git","2026-05-15",1024,"typography-portrait-art-325",[19,28,35,43,52,60],{"id":20,"url":21,"thumb":21,"prompt":22,"prompt_cn":8,"prompt_en":22,"tags":23,"title":26,"source_repo":14,"date":15,"width":16,"height":16,"slug":27,"urlWebp":21,"thumbWebp":21},258,"\u002Fimage\u002Fed71e1c90a5e4bec.webp","[中文]\n一幅细节丰富的工笔画，描绘了一位唐朝贵族女子在御花园中漫步。她看起来优雅而平静。\n\n她手里拿着一根金色的牵引绳。牵引绳的尽头是一只可怕的**异形怪物（出自电影《异形》）**。然而，这只异形穿着一件**可爱的粉色丝绸马甲**，并且表现得像一只训练有素的狗。\n\n背景有牡丹和蝴蝶。\n\n**在右下角，有一个红色的竖排艺术家印章，写着“吴先生”（Mr. Wu），风格像水印一样。** --ar 3:4\n\n[English]\nA finely detailed Gongbi painting of a noble Tang Dynasty lady taking a stroll in the imperial garden. She looks elegant and calm.\n\nShe is holding a gold leash. At the end of the leash is a terrifying **Xenomorph monster (from the movie Alien)**. However, the Xenomorph is wearing a **cute pink silk vest** and is behaving like a well-trained dog.\n\nBackground features peonies and butterflies.\n\n**In the bottom right corner, a single red vertical artist chop seal reads \"吴先生\" (Mr. Wu), stylized like a watermark.** --ar 3:4",[24,25,12],"Animal","Landscape","Tang Lady Walking Alien Monster Gongbi Art","animal-landscape-art-258",{"id":29,"url":30,"thumb":30,"prompt":31,"prompt_cn":8,"prompt_en":31,"tags":32,"title":33,"source_repo":14,"date":15,"width":16,"height":16,"slug":34,"urlWebp":30,"thumbWebp":30},268,"\u002Fimage\u002F1c060557692f4b31.webp","{argument name=\"voice\" default=\"chatgpt voice\"} if it were a character",[11],"Voice Character Portrait Illustration","portrait-268",{"id":36,"url":37,"thumb":37,"prompt":38,"prompt_cn":8,"prompt_en":38,"tags":39,"title":41,"source_repo":14,"date":15,"width":16,"height":16,"slug":42,"urlWebp":37,"thumbWebp":37},69,"\u002Fimage\u002F8685965806b344e4.webp","Transform the uploaded logo into a hyper-realistic scene where the logo silhouette is formed by iron filings reacting to a magnetic field. The logo must keep its exact shape and proportions, appearing as if a magnet shaped like the logo (or hidden beneath the surface) is influencing the filings to align naturally into that recognizable formation. Material details:\nFine iron filings with sharp, spiky, needle-like structures.\nDense clustering along magnetic field lines forming the logo silhouette.\nVariation in density — thick near magnetic poles, thinner outward.\nMatte dark metallic texture with subtle reflections.\n\nPhysics behavior:\nIron filings must follow realistic magnetic field patterns — radial and curved lines forming spikes and ridges.\nStronger attraction zones create thicker, raised clusters.\nOuter areas show softer, more dispersed alignment.\nNatural randomness and slight irregularity — no perfect edges.\nSome loose filings scattered beyond the main shape.\n\nSurface interaction:\nFlat surface such as a lab table, glass plate, or matte black tray.\nFilings resting on surface but visibly lifted in areas due to magnetic force (spiky texture).\nSubtle dust and micro particles around.\n\nEnvironment & human presence:\nRealistic classroom, science lab, or creative studio environment.\nA person partially visible — hands holding or moving a magnet beneath the surface or nearby.\nPossibly a child or adult observing or interacting (adds emotional curiosity).\nOther subtle elements: notebooks, tools, or lab items out of focus.\n\nLighting:\nDirectional overhead light creating shadows from raised filings.\nSubtle highlights on metallic edges.\nBalanced natural or indoor lighting.\n\nAtmosphere:\nCuriosity.\nDiscovery.\nEducational yet visually satisfying.\nQuiet but engaging moment.\n\nCamera & composition:\nTop-down or slightly angled close-up view.\nLogo clearly visible through iron filing formation.\nHuman hands or interaction slightly off-center for storytelling.\n\nFormat:\nAspect ratio: STRICT 4:5 vertical.\nNo text overlays.\n\nStyle:\nHyper-real macro + environmental photography.\nPhysically accurate magnetic behavior.\nCinematic yet grounded realism.",[10,40,11],"Photography","Hyper-Realistic Magnetic Iron Filings Logo Art","typography-photography-portrait-69",{"id":44,"url":45,"thumb":45,"prompt":46,"prompt_cn":8,"prompt_en":46,"tags":47,"title":49,"source_repo":14,"date":50,"width":16,"height":16,"slug":51,"urlWebp":45,"thumbWebp":45},2251,"\u002Fimage\u002F2a45a4c0c1cd4fdd.webp","《导览式科普绘本》提示词：\n\n请根据【主题】创作一张高完成度的「导览式科普绘本」风格插画。\n\n这是一张结合“大型场景主视觉 + 导览路线 + 可爱导览 IP + 知识站点 + 儿童科普绘本质感”的场景导览式科普图解页。画面需要让观者像被带着参观一个复杂系统一样，边看边理解主题背后的运行逻辑、空间结构、流程关系和关键知识点。\n\n【基础设定】\n主题：【填写主题，例如发射场的一天 \u002F 一个集装箱的旅行 \u002F 地铁站里的秘密路线 \u002F 下潜到深海的一小时 \u002F 机场如何运转 \u002F 医院急诊系统 \u002F 智慧农场 \u002F 消防站出警流程】\n画幅比例：【4:3 横版】\n主色调：【根据主题自动匹配，整体保持明亮、清爽、儿童友好】\n风格方向：【现代儿童科普绘本 \u002F 场景导览式图解 \u002F 高完成度数字插画】\n\n【核心表达】\n请围绕【主题】设计一个完整的大型场景或复杂系统。画面中必须有一个明确的主视觉场景，例如大型设施、交通系统、科技装备、自然探索场景、城市公共系统或生产流程。主体要足够清晰、有规模感、有细节，能够成为第一眼的视觉中心。\n\n画面要通过“导览路线”的方式组织信息，避免仅把场景做成静态陈列。请设计一条清晰的参观路线、流程路线、时间线或空间动线，让读者可以沿着路线一步步理解这个系统是如何运行的。\n\n【导览 IP 设计】\n请为本图设计一个原创、可爱、亲和的导览小 IP。导览 IP 可以是小动物、小朋友、拟人化工具或其他适合主题的原创形象，但必须具有独立原创性，不要照搬任何参考图中的角色、动物形象、服装、配色或搭档关系。\n\n导览 IP 的作用是：\n1. 开场介绍主题\n2. 指向关键知识点\n3. 引导读者顺着路线阅读\n4. 增加儿童绘本的陪伴感和趣味性\n\n导览 IP 可以在画面中出现 2-3 次，但不要过度抢主视觉。角色应圆润、可爱、有表情、有动作，适合儿童科普绘本。\n\n【信息结构】\n画面中请设置 3-6 个“知识站点”，每个站点用简短中文标签和短说明表达。站点命名可以采用：\n- 第1站｜xxx\n- 第2站｜xxx\n- 重点观察｜xxx\n- 小知识｜xxx\n- 为什么｜xxx\n- 如何工作｜xxx\n\n每个知识点都要围绕主题的核心运行逻辑展开，不要写空泛说明。文字要短、清楚、自然，避免长段落，适合儿童阅读。\n\n【画面模块】\n整张图建议包含以下模块：\n1. 顶部主标题区：清楚写出主题名称\n2. 开场导览区：导览 IP 引出主题\n3. 大型主场景区：展示主题系统的完整场景\n4. 导览路线区：用箭头、虚线、路径、时间节点或流程线串联知识点\n5. 知识站点区：用小信息框、导览牌、局部标注展示关键知识\n6. 小百科 \u002F 小贴士区：补充一个有趣知识\n7. 收尾区：让导览 IP 做简短总结或引导\n\n【构图要求】\n画面采用 4:3 横版构图，整体像一本高质量儿童科普绘本的跨页，也像一张儿童科技馆导览图。画面需要有清晰的视觉重心：大型主场景占据主要空间，导览路线贯穿画面，知识模块自然分布在周围。信息丰富但不能杂乱，阅读路径要顺畅。\n\n【视觉风格】\n整体采用现代儿童科普绘本风格：\n- 明亮、清爽、干净的色彩\n- 清晰自然的手绘线条\n- 高完成度数字插画质感\n- 细节丰富但有秩序\n- 可爱但不低幼\n- 有科普图解感\n- 有导览地图感\n- 场景真实可信，但表达方式亲和\n\n【文字与标注】\n文字以中文为主，使用短标题、短标签、简短说明。不要生成大段复杂文字。信息框应像儿童科普书中的导览牌、知识卡片或小贴士。文字要尽量清晰、简洁、可读。\n\n【最终目标】\n让整张图像一页高质量的儿童科普绘本：孩子第一眼被可爱角色和大场景吸引，第二眼能顺着路线读懂系统如何运行，第三眼还能继续发现细节和知识点。画面要具有系列化潜力，方便后续替换不同主题继续创作同类型图片。",[11,48,10],"Education","Engaging Children's Science Guide Poster","2026-05-26","portrait-education-typography-2251",{"id":53,"url":54,"thumb":54,"prompt":55,"prompt_cn":8,"prompt_en":55,"tags":56,"title":58,"source_repo":14,"date":15,"width":16,"height":16,"slug":59,"urlWebp":54,"thumbWebp":54},311,"\u002Fimage\u002F1cc9c440d7804142.webp","{\n  \"type\": \"anime movie production pitch document\",\n  \"overall_layout\": \"split layout with a large cinematic movie poster on the top half and a grid of 5 detailed reference sheets on the bottom half\",\n  \"top_section\": {\n    \"type\": \"movie poster\",\n    \"visual\": \"A man, a woman, and a dog standing on a ruined city street, facing away from the viewer, looking towards a colossal, porous, web-like alien structure dominating the sky. A rusty 'RESTRICTED AREA' sign is on the right.\",\n    \"typography\": {\n      \"title\": \"{argument name=\\\"movie title\\\" default=\\\"劇場版 巨骸の向こう側 Fallen Colossus\\\"}\",\n      \"release_date\": \"{argument name=\\\"release date\\\" default=\\\"2027.11.28 ROADSHOW\\\"}\",\n      \"tagline\": \"そこにあるのは、まだ「説明」されていないもの。\",\n      \"credits_studio\": \"{argument name=\\\"studio name\\\" default=\\\"WIT STUDIO\\\"}\"\n    }\n  },\n  \"bottom_sections\": [\n    {\n      \"title\": \"{argument name=\\\"male character name\\\" default=\\\"来栖 武 \u002F Kurusu Takeru\\\"}\",\n      \"type\": \"character reference sheet\",\n      \"elements\": {\n        \"full_body_poses\": 3,\n        \"expressions\": 3,\n        \"detail_shots\": 8,\n        \"description\": \"Male protagonist in dark tactical jacket and cargo pants. Includes front, back, and side full-body views, headshots, and detailed callouts for gloves, boots, backpack, and radio.\"\n      }\n    },\n    {\n      \"title\": \"{argument name=\\\"female character name\\\" default=\\\"大城 真那 \u002F Oshiro Mana\\\"}\",\n      \"type\": \"character reference sheet\",\n      \"elements\": {\n        \"full_body_poses\": 3,\n        \"expressions\": 3,\n        \"detail_shots\": 6,\n        \"description\": \"Female protagonist in grey tactical uniform. Includes front, back, and side full-body views, headshots, and detailed callouts for jacket, boots, ID badge, and pouch.\"\n      }\n    },\n    {\n      \"title\": \"カゲ (Kage) 設定画\",\n      \"type\": \"animal character reference sheet\",\n      \"elements\": {\n        \"full_body_poses\": 4,\n        \"expressions\": 4,\n        \"detail_shots\": 5,\n        \"description\": \"Dog companion. Includes side, front, back, and angled full-body views, headshots, and detailed callouts for fur texture, paws, and a motorcycle sidecar.\"\n      }\n    },\n    {\n      \"title\": \"第7巨骸 (Remnant-7) 内部区画 設定画\",\n      \"type\": \"environment and vehicle reference sheet\",\n      \"elements\": {\n        \"large_diagrams\": 1,\n        \"environment_thumbnails\": 4,\n        \"vehicle_designs\": 1,\n        \"description\": \"Cross-section of the porous alien structure, smaller environment thumbnails, and a motorcycle design featuring the characters.\"\n      }\n    },\n    {\n      \"title\": \"Concept Art\",\n      \"type\": \"scene illustration\",\n      \"elements\": {\n        \"characters\": 3,\n        \"vehicles\": 1,\n        \"description\": \"The male character, female character, and dog with a motorcycle sidecar parked in front of the glowing, porous alien structure.\"\n      }\n    }\n  ]\n}",[11,57,24],"Storyboard","Fallen Colossus Anime Movie Pitch Deck","portrait-storyboard-animal-311",{"id":61,"url":62,"thumb":62,"prompt":63,"prompt_cn":8,"prompt_en":63,"tags":64,"title":66,"source_repo":14,"date":15,"width":16,"height":16,"slug":67,"urlWebp":62,"thumbWebp":62},352,"\u002Fimage\u002Fa46d63d65ec54c14.webp","[CORE TASK]\nTransform the provided input image into a pose-and-light analysis sheet.\n\nThis is NOT a finished character illustration.\nThis is NOT a clothing sheet.\nThis is NOT a beauty-preserving redraw.\n\nThis is a white-line rough mannequin conversion.\n\n[PRIMARY GOAL]\nExtract and visualize only:\n- pose structure\n- body balance\n- camera angle\n- body line flow\n- inferred light source placement\n- illuminated areas and light intensity\n\n[INPUT ROLE]\nUse the provided image as the strict anchor for:\n- pose\n- camera angle\n- body tilt\n- weight distribution\n- approximate lighting situation\n\nDo NOT preserve:\n- face rendering\n- hairstyle rendering\n- clothing detail\n- accessories\n- weapon detail\n- background architecture\n- character identity\n- emotional expression\n\n[FIGURE CONVERSION]\nsingle rough mannequin-like human figure\nwhite body contour lines\nwhite internal construction lines\nsimple mannequin head\nno face\nno eyes\nno mouth\nno eyelashes\nno personality\nno individual identity\n\nhuman figure should look like:\n- rough pose mannequin\n- anatomy proxy\n- line-based body guide\n- structural sketch\n- white-line rough dummy\n\nkeep:\n- pose readability\n- silhouette flow\n- head tilt\n- torso direction\n- pelvis direction\n- limb placement\n\n[BACKGROUND]\npure black background\nnegative-style dark field\nno scenery\nno props\nno architecture\nno environmental storytelling\n\n[LINE STYLE]\nrough white line drawing\nclean but sketch-like\nconstruction-line feeling\nanatomy guide lines visible\njoint flow visible\nbody contour emphasized\nno polished illustration finish\n\n[LIGHT ESTIMATION]\npredict the likely light source positions from the input image\nvisualize the light sources and illuminated areas using green glow only\n\nuse green light intensity with variation:\n- strongest green where the light directly hits\n- medium green for wrap light\n- soft green for reflected or fading light\n\nmark the estimated light sources with labels and arrows such as:\n- Main Light\n- Rim Light\n- Fill Light\n- Floor Bounce\n- Back Light\nonly if appropriate\n\nIMPORTANT:\ndo not invent random lights\ninfer lighting from the original input image\nif the lighting is ambiguous, keep the annotations simple and plausible\n\n[GREEN LIGHT VISUALIZATION]\nshow green glow on:\n- head \u002F skull plane\n- neck\n- shoulders\n- chest plane\n- ribcage direction\n- pelvis edge\n- thigh planes\n- knee contact points\n- floor contact bounce if applicable\n\nuse green light not as decoration,\nbut as lighting analysis information\n\n[POSE PRIORITY]\n1. preserve pose structure\n2. preserve camera angle\n3. preserve body balance\n4. preserve head-torso relationship\n5. visualize likely light direction\n6. show illuminated areas with readable green intensity variation\n\n[NEGATIVE]\nfinished person,\ncute girl,\ndetailed face,\nhair rendering,\nclothing rendering,\nweapon emphasis,\nbeautiful anatomy",[11,12,65],"Product","Mannequin Pose Analysis with Green Light","portrait-art-product-352",1781516702346]