主体:

subject: core_identity: "" # Who or what is the main subject (character name, creature type, object) physical_appearance: age_gender_presentation: "" ethnicity_skin_tone: "" face: face_shape: "" facial_features: "" eyes: "" eyebrows: "" nose: "" lips_mouth: "" makeup: "" hair: hairstyle: "" hair_color: "" hair_length_texture: "" hair_details: "" body: body_type: "" height_proportions: "" skin_details: "" muscle_fat_distribution: "" clothing_attire: overall_style: "" upper_body: base_layer: "" outer_layer: "" details: "" lower_body: base_layer: "" outer_layer: "" details: "" footwear: "" accessories: head_accessories: "" neck_accessories: "" hand_accessories: "" body_accessories: "" special_items: "" pose_expression: body_pose: "" hand_gesture: "" facial_expression: "" eye_direction_gaze: "" overall_demeanor: "" scene_environment: primary_setting: "" time_of_day: "" weather_atmosphere: "" background: main_background_elements: "" depth_layers: "" environmental_details: "" foreground_elements: "" props_objects: "" particles_effects: "" artistic_direction: art_medium: "" # photorealistic, digital painting, oil on canvas, 3D render, etc. rendering_style: "" # hyperrealistic, painterly, illustrative, cinematic, etc. aesthetic_vibe: "" # cyberpunk, dark fantasy, solarpunk, etc. color_palette_mood: "" dominant_colors: "" artist_references: "" # "in the style of [artist1], [artist2]" composition_influences: "" cinematography: composition: shot_type: "" # extreme close-up, close-up, medium shot, full body, wide shot framing: "" # centered, rule of thirds, symmetrical, leading lines camera_angle: "" # eye level, low angle, high angle, dutch tilt, over-the-shoulder perspective: "" lighting: light_source: "" lighting_direction: "" lighting_quality: "" # soft, hard, volumetric, cinematic, neon, natural key_light: "" fill_light: "" rim_light_backlight: "" color_temperature: "" atmosphere_mood: overall_mood: "" emotional_tone: "" cinematic_feel: "" technical_quality: detail_level: "extremely detailed, intricate details, fine details" sharpness_clarity: "sharp focus, crisp details, high resolution" texture_quality: "highly detailed textures, realistic material rendering" quality_boosters: "masterpiece, best quality, ultra-detailed, 8k uhd, professional illustration, award-winning" anatomy_technical: "perfect anatomy, beautiful detailed face, detailed hands, detailed fingers" render_quality: "octane render, unreal engine 5, ray tracing, global illumination" # use only when appropriate # ============================================================ # NEGATIVE PROMPT STRUCTURE (Deeply Nested) # ============================================================ negative_prompt: anatomy_issues: - "deformed, bad anatomy, extra limbs, missing limbs, fused fingers, too many fingers" - "mutated hands, poorly drawn hands, mutated face" quality_issues: - "low quality, worst quality, blurry, out of focus, low resolution" - "jpeg artifacts, compression artifacts, noise, grain" unwanted_elements: - "watermark, text, signature, logo, username" - "censored, bar censor, mosaic censor" - "extra objects, cluttered background" style_issues: - "cartoon, 3d render, cgi, plastic, doll-like" # add/remove depending on desired style general: - "overexposed, underexposed, washed out colors" # ============================================================ # ASSEMBLY INSTRUCTIONS (How top users actually use this) # ============================================================ assembly_guide: recommended_order: | subject.physical_appearance + https://t.co/WYCxCaos8B_attire + subject.pose_expression → scene_environment → artistic_direction → cinematography → technical_quality.quality_boosters (always at the end) tips_for_max_quality: - Be extremely specific in the subject section (this is where most quality comes from in 2026 models) - Use the nested clothing layers for complex outfits - Keep lighting and cinematography sections very detailed — modern models respond extremely well to cinematic language - Always put quality boosters at the very end of the positive prompt - Customize the negative prompt based on the art style you're going for - You can leave some fields empty if not relevant example_filled_snippet: | A highly detailed portrait of a 24 year old Japanese woman with sharp facial features, long straight black hair with subtle blue highlights, wearing a detailed black cyberpunk jacket over a white techwear shirt, standing in a rainy neon Tokyo street at night, cinematic lighting, dramatic rim lighting, moody atmosphere, masterpiece, best quality, ultra-detailed, sharp focus, 8k

主体: 核心身份: "" # 主要主体是谁或是什么（角色名、生物种类、物体） 外貌特征: 年龄与性别呈现: "" 民族与肤色: "" 面部: 脸型: "" 面部特征: "" 眼睛: "" 眉毛: "" 鼻子: "" 嘴唇与嘴巴: "" 妆容: "" 头发: 发型: "" 发色: "" 发长与质地: "" 头发细节: "" 身体: 体型: "" 身高与比例: "" 皮肤细节: "" 肌肉与脂肪分布: "" 服装服饰: 整体风格: "" 上身: 内层: "" 外层: "" 细节: "" 下身: 内层: "" 外层: "" 细节: "" 鞋履: "" 配饰: 头部配饰: "" 颈部配饰: "" 手部配饰: "" 身体配饰: "" 特殊物品: "" 姿态与表情: 身体姿态: "" 手势: "" 面部表情: "" 视线方向与注视: "" 整体气质: "" 场景环境: 主要场景: "" 时间: "" 天气与氛围: "" 背景: 主要背景元素: "" 景深层次: "" 环境细节: "" 前景元素: "" 道具与物件: "" 粒子与特效: "" 艺术指导: 艺术媒介: "" # 照片级真实、数字绘画、布面油画、3D 渲染等 渲染风格: "" # 超写实、绘画风格、插画风格、电影感等 美学基调: "" # 赛博朋克、黑暗奇幻、太阳能朋克等 配色与情绪: "" 主导颜色: "" 艺术家参考: "" # "in the style of [artist1], [artist2]" 构图影响: "" 电影摄影: 构图: 景别: "" # 特写、大特写、中景、全身、远景 取景: "" # 居中、三分法、对称、引导线 拍摄角度: "" # 平视、仰视、俯视、荷兰角、过肩 透视: "" 光照: 光源: "" 光照方向: "" 光照质感: "" # 柔和、硬光、体积光、电影感、霓虹、自然 主光: "" 补光: "" 轮廓光与逆光: "" 色温: "" 氛围与情绪: 整体情绪: "" 情感基调: "" 电影感: "" 技术质量: 细节层次: "extremely detailed, intricate details, fine details" 锐度与清晰度: "sharp focus, crisp details, high resolution" 纹理质量: "highly detailed textures, realistic material rendering" 质量增强词: "masterpiece, best quality, ultra-detailed, 8k uhd, professional illustration, award-winning" 解剖与技术: "perfect anatomy, beautiful detailed face, detailed hands, detailed fingers" 渲染质量: "octane render, unreal engine 5, ray tracing, global illumination" # 仅在合适时使用 # ============================================================ # 负面提示结构（深度嵌套） # ============================================================ negative_prompt: 解剖问题: - "deformed, bad anatomy, extra limbs, missing limbs, fused fingers, too many fingers" - "mutated hands, poorly drawn hands, mutated face" 质量问题: - "low quality, worst quality, blurry, out of focus, low resolution" - "jpeg artifacts, compression artifacts, noise, grain" 不需要的元素: - "watermark, text, signature, logo, username" - "censored, bar censor, mosaic censor" - "extra objects, cluttered background" 风格问题: - "cartoon, 3d render, cgi, plastic, doll-like" # 根据目标风格增删 通用: - "overexposed, underexposed, washed out colors" # ============================================================ # 组装指南（顶级用户实际用法） # ============================================================ assembly_guide: recommended_order: | subject.physical_appearance + https://t.co/WYCxCaos8B_attire + subject.pose_expression → scene_environment → artistic_direction → cinematography → technical_quality.quality_boosters (always at the end) tips_for_max_quality: - Be extremely specific in the subject section (this is where most quality comes from in 2026 models) - Use the nested clothing layers for complex outfits - Keep lighting and cinematography sections very detailed — modern models respond extremely well to cinematic language - Always put quality boosters at the very end of the positive prompt - Customize the negative prompt based on the art style you're going for - You can leave some fields empty if not relevant example_filled_snippet: | A highly detailed portrait of a 24 year old Japanese woman with sharp facial features, long straight black hair with subtle blue highlights, wearing a detailed black cyberpunk jacket over a white techwear shirt, standing in a rainy neon Tokyo street at night, cinematic lighting, dramatic rim lighting, moody atmosphere, masterpiece, best quality, ultra-detailed, sharp focus, 8k