← 返回首页
技能

Ai Image Prompts

tencent_skillhub · 1.0.9 · 下载量 1,827 · 热度分 55.6

产品介绍

从包含10,000+真实图像生成提示的精选库中推荐提示。兼容任何AI图像模型,包括Nano Banana Pro、Nano Banana 2、Seedream等。

资源信息

作者:dophinl
分类:图文音视频多模态生成
版本:1.0.9
认证方式: 无需认证
是否付费: 官方未说明

许可证信息摘自官方源,可能过时或缺失。商用前请以原仓库的 LICENSE 文件为准。

标签
图片生成

提示词库 · 2.3万

数据源:社区提示词库(MIT 许可,每日同步)

共 15141 条提示词,按 11 个用途归类;同一条可归入多个用途,因此分类条数相加会大于去重后的总数。

命中 725 条

Cyberpunk City Block Collectible Diorama

game-asset

A detailed prompt for generating a stylized 3D miniature diorama of a layered cyberpunk city block, combining historic ground structures with neon skyscrapers. The style is specified as a premium collectible, sculpted from smooth ceramic or resin with clean, toy-like proportions and soft studio product lighting.

展开提示词全文
SCENE:
A layered cyberpunk city block combining low historic structures at ground level with towering neon skyscrapers rising above, connected by bridges, cables, and stacked infrastructure.
STYLE:
Render as a premium stylized 3D collectible miniature diorama on a fully visible, centered self-contained display base, shown from a slightly elevated 3/4 tabletop angle against a clean neutral studio gradient background. Maintain collectible figurine scale and clean display presentation.
All elements should appear sculpted from smooth matte ceramic, painted resin, or stylized polymer with rounded edges, simplified geometry, and slightly exaggerated toy-like proportions. Keep surfaces clean and intentionally stylized. Avoid fingerprints, rough handmade textures, workshop scenes, macro photography depth-of-field, or photoreal clay imperfections.
Use soft professional studio product lighting: gentle global illumination, warm key light, soft ambient bounce, controlled shadows, and optional warm interior window glow.
Render any water or translucent materials as poured clear resin with smooth transparency, subtle reflections, soft internal depth, and slightly thick stylized edges.
Include small environmental details but keep everything cohesive, simplified, and sculptural. Prioritize clean stylized modeling over photorealistic texture detail.
Overall intent: premium collectible miniature, stylized 3D marketplace presentation, cozy whimsical mood, smooth sculpt clarity, controlled studio rendering, avoid real-world photography aesthetics

Generate Ghostbusters Scene with Gozer and Stay Puft

game-asset

An image generation prompt for Nano Banana Pro (Gemini 3 Pro Image) requesting a scene from Ghostbusters featuring the text 'Ghostbusters' at the bottom, and incorporating the destructive god Gozer and the Stay Puft Marshmallow Man in a good composition.

展开提示词全文
Ghostbusters
'Ghostbusters' at the bottom of the screen
Also,
Include the destructive god Gozer and the Stay Puft Marshmallow Man
in a good composition!

Ancient Egyptian Ritual with Alien Entity Prompt

game-asset

A detailed English prompt used with Nano Banana Pro to generate an image, which was then used to create a video in Higgsfield Cinema Studio 2.0. The prompt describes a monochrome, ritualistic scene in ancient Egypt involving a ceremonial procession, sacred geometry, and a gigantic, unsettling alien entity presiding over the scene.

展开提示词全文
Monochrome ritualistic scene set in ancient Egypt, a ceremonial procession of ancient Egyptian people moving a massive square stone block as part of a sacred rite, bodies arranged in deliberate patterns, synchronized movements, sand disturbed like ritual markings. A gigantic alien entity presides over the scene like an ancient god. Its skin is thick and cracked like elephant hide, aged and primordial; its legs are impossibly thin and unnaturally long like a deer’s, elevating its body into an inhuman silhouette. The face is deeply unsettling, elongated and expressionless, evoking Anunnaki-like beings and forgotten star gods. The alien radiates silent authority, worshipped rather than feared. High-contrast lighting, deep shadows, sacred geometry in composition, mythological atmosphere, surreal historical ritual, solemn and ominous mood, timeless and forbidden

Ancient Jungle Temple Diorama Collectible Prompt

game-asset

A detailed prompt for generating an image of an ancient jungle temple complex diorama. It specifies the scene layout (stepped temple, courtyards, moss, vines) and the required style: a premium stylized 3D collectible miniature, sculpted from smooth matte ceramic/resin, with controlled studio lighting and a cozy, whimsical mood.

展开提示词全文
SCENE:
A larger ancient jungle temple complex featuring a stepped main temple, several surrounding shrine structures, broken ceremonial courtyards, moss-covered staircases, carved statues, fragmented stone walls, and a shallow reflecting pool, all surrounded by dense tropical foliage and vines. Arrange the ruins so the space feels like a compact, layered miniature district with multiple clustered architectural elements, varied heights, and interconnected pathways, filling the platform naturally without empty gaps and reading clearly as a contained archaeological jungle ruin diorama.
STYLE:
Render as a premium stylized 3D collectible miniature diorama on a fully visible, centered self-contained display base, shown from a slightly elevated 3/4 tabletop angle against a clean neutral studio gradient background. Maintain collectible figurine scale and clean display presentation.
All elements should appear sculpted from smooth matte ceramic, painted resin, or stylized polymer with rounded edges, simplified geometry, and slightly exaggerated toy-like proportions. Keep surfaces clean and intentionally stylized. Avoid fingerprints, rough handmade textures, workshop scenes, macro photography depth-of-field, or photoreal clay imperfections.
Use soft professional studio product lighting: gentle global illumination, warm key light, soft ambient bounce, controlled shadows, and optional warm interior window glow.
Render any water or translucent materials as poured clear resin with smooth transparency, subtle reflections, soft internal depth, and slightly thick stylized edges.
Include small environmental details but keep everything cohesive, simplified, and sculptural. Prioritize clean stylized modeling over photorealistic texture detail.
Overall intent: premium collectible miniature, stylized 3D marketplace presentation, cozy whimsical mood, smooth sculpt clarity, controlled studio rendering, avoid real-world photography aesthetics.

Miniature Alpine Mountainside Village Scene

game-asset

A prompt for generating a detailed scene of a miniature alpine mountainside village, specifying the arrangement of wooden chalets, snow-covered paths, glowing window lights, and surrounding elements like pine trees and a frozen pond.

展开提示词全文
SCENE:
A larger alpine mountainside village composed of multiple wooden chalet lodges of varying sizes arranged along winding snow-covered paths, including a central lodge hall, several smaller cabins, storage sheds, fenced yards, stacked firewood zones, snow-covered pine trees, and a small frozen pond connected by stone walkways. Warm window lights glow across the settlement while snow rests on rooftops and railings. Arrange t

Cinematic Concept Art for Ancient Astronomical Observatory

game-asset

A structured prompt for generating high-end cinematic concept art of an ancient astronomical observatory carved into a towering cliff face, emphasizing monumental scale, early morning directional sunlight, and a mood of historical mystery, suitable for AAA game environments.

展开提示词全文
SHORT STRUCTURED PROMPT — CINEMATIC OBSERVATORY TEST
STYLE
High-end cinematic stylized-realism digital concept art, AAA environment
keyframe illustration, film matte-painting depth construction, professional
game environment rendering, minimal line art, value-driven form separation,
painted edges, controlled edge hierarchy, large-to-small mass construction,
clean gradient digital color blending, hybrid hand-painted procedural texture,
macro readability prioritized with selective micro-detail in focal structures,
studio-quality composition, polished senior-level production finish.
SUBJECT
Ancient astronomical observatory carved into a towering cliff face, featuring
layered stone platforms, carved stairways, suspended bridges, and a monumental
circular astronomical device embedded in the rock wall.
ENVIRONMENT
Vast mountainous frontier with deep forest valleys, winding river systems,
and small distant settlements arranged with monumental geological scale and
clear large-shape composition.
LIGHTING
Early morning directional sunlight defining major planes clearly, gentle
atmospheric depth haze in distant layers while keeping strong foreground
clarity.
CAMERA
Wide cinematic establishing shot from an elevated hillside vantage point with
natural foreground framing, strong midground structure emphasis, and expansive
layered background depth.
MOOD
Grand historical mystery and discovery, like a major story-location reveal
from a high-budget fantasy film or AAA RPG cinematic.
DETAIL CONTROL
Professional focal-detail distribution, strong readable primary masses first,
secondary structures next, fine detail only in anchor regions, avoid uniform
micro-detail everywhere.
QUALITY TARGET
Ultra-clean studio-grade cinematic environment illustration, portfolio-level
concept art readability both at full view and zoom.

Cinematic Concept Art for Ancient Abandoned Jungle City

game-asset

A highly structured master prompt designed for generating high-end cinematic stylized-realism concept art of an ancient, abandoned jungle city plaza, focusing on deep environmental layering, early morning lighting filtering through the canopy, and a sense of ancient discovery.

展开提示词全文
ONE-CLICK GENERIC MASTER PROMPT — ANCIENT JUNGLE MORNING
STYLE
High-end cinematic stylized-realism digital concept art, AAA environment
keyframe illustration, film matte-painting depth construction, professional
game environment rendering, minimal line art, value-driven form separation,
painted edges, controlled edge hierarchy, large-to-small mass construction,
clean gradient digital color blending, hybrid hand-painted procedural texture,
macro readability prioritized with selective micro-detail in focal structures,
studio-quality composition, polished senior-level production finish.
SUBJECT / SCENE
In an Ancient Jungle Morning Setting, a long-abandoned jungle city plaza with cracked statues and vine-covered arches.
ENVIRONMENT
Primeval overgrown jungle landscape with monumental ancient natural formations,
dense layered vegetation, colossal trees, tangled roots, mist-filled clearings,
hidden stone remnants or forgotten pathways partially reclaimed by nature.
Large readable terrain masses first, with deep environmental layering and
strong spatial separation between foreground foliage, mid jungle structures,
and distant canopy silhouettes.
LIGHTING
Fresh early-morning jungle illumination with low-angle sunlight filtering
through dense canopy gaps in structured directional shafts. Light should
clearly define major structural planes while allowing gentle atmospheric mist
to separate depth layers and enhance spatial readability without obscuring
primary forms.
CAMERA
Cinematic exploration-style wide establishing shot positioned slightly above
ground level as if the viewer has just discovered the scene. Use natural
foreground framing through vegetation or terrain forms, clear midground focal
anchor, and deep layered jungle distance to reinforce cinematic matte-painting
depth construction.
MOOD
Sense of ancient discovery, untouched history, and quiet natural majesty.
The scene should feel like the first reveal of a lost jungle location in a
high-budget adventure film or AAA exploration game.
DETAIL CONTROL
Maintain professional focal-detail hierarchy. Preserve large readable jungle
masses first, secondary structural forms next, and apply fine detail only in
primary focal regions. Avoid uniform high-frequency foliage noise or excessive
micro-detail in distant canopy layers.
QUALITY TARGET
Ultra-clean studio-grade cinematic jungle environment illustration, portfolio-
level concept art readability, strong depth hierarchy, and polished production-
ready rendering quality.

Massive Truck Launch and Landing Video Prompt

game-asset

A detailed motion prompt for generating a cinematic video scene where a massive black RAM TRX-style truck launches aggressively over a man, landing with an explosion of snow and stopping inches from his back, emphasizing realistic proportions and heavy suspension compression.

展开提示词全文
Massive black RAM TRX–style truck realistically launches upward behind a man (as if there is an invisible ramp hidden just outside the camera frame) and aggressively lands with a powerful explosion of snow.
The truck’s suspension compresses heavily as it hits the ground, drives forward a few meters, and stops just inches from the man’s back. The vehicle maintains true-to-life proportions with no distortion.
The man remains completely still. The shot ends with the truck standing on the ground directly behind him.

Wizard in a Box Miniature Diorama

game-asset

A highly detailed prompt for generating an ultra-realistic, cinematic fantasy miniature diorama. The core concept is a two-level wizard library tower where the architecture is seamlessly integrated into an ornate open antique box, featuring a working wizard, floating books, and dramatic lighting.

展开提示词全文
SUBJECT: A two-level miniature wizard library tower where the ornate open antique box itself forms the architecture, the base container acting as the full ground floor chamber with the box walls serving directly as the structural stone or wooden walls, featuring an enthusiastic wizard actively working at a magical desk with floating books and glowing runes, the scene fully integrated into the box structure and not a separate room placed inside, not a wall frame or display frame
ENVIRONMENT: The base of the ornate box is the ground floor surface and foundation, its interior sides transformed into the ground-floor walls containing carved doorways and window openings that are cut seamlessly into the box sides, perfectly aligned with and blending into the exterior ornamental carvings, a mezzanine first floor built above with railings and shelves attached directly to the box walls, connected by a spiral staircase, the inside of the open lid acting as the vaulted tower ceiling with faint arcane symbols
DETAILS: Intricate baroque or gothic carved antique box with engraved panels, hinges clearly visible, exterior ornamentation continuing logically into interior architecture, doors and windows precisely cut into the box sides following the decorative patterns, believable miniature structural joins where beams meet the container walls, ultra-fine textures on books, wood grain, stone blocks, magical glyphs, levitating books, parchment pages, and drifting magical particles surrounding the wizard
STYLE: Ultra-realistic cinematic fantasy miniature diorama photography, museum-grade handcrafted realism, extreme macro detail, physically believable materials, strong architectural integration of container and structure, Nano Banana Pro optimized
LIGHTING: Warm magical glow from the wizard’s active spell combined with a soft overhead shaft of cool light entering from above, subtle rays catching window openings in the box walls, volumetric dust and glowing particle trails enhancing depth and scale
COLOR: Deep mahogany shelves, aged parchment creams, warm amber magical light, jewel-toned magical blues and violets, muted stone greys, antique bronze or dark carved wood tones of the ornate box fully preserved
CAMERA: Front-facing slightly elevated macro shot looking into the open box so the integrated ground floor walls, cut door openings, mezzanine, spiral staircase, and working wizard are all visible clearly, shallow depth of field centered on the wizard and structural integration
ASPECT RATIO: 4:3

Worm's-Eye Perspective Character Shot (Likeness Preservation)

game-asset

A detailed prompt for generating a cinematic, worm's-eye perspective shot of a character (Beast/creature), focusing on extreme scale distortion using an ultra-wide lens, strict preservation of the reference image's likeness and materials, and dramatic upward rim lighting.

展开提示词全文
A cinematic worm’s-eye perspective shot of the character captured from ground level, camera almost touching the surface beneath them. Ultra-wide 14–18mm lens. The camera is angled sharply upward, making the character tower massively in frame.

Likeness Preservation:

Maintain exact proportions, eye shape, stylization, and material identity from the reference. Do not humanize or reinterpret anatomy. Perspective exaggeration must come only from camera placement, not redesign.

Composition:

The lowest part of the character dominates the foreground due to proximity to the lens. Boots, feet, lower garment, or base form appear large and imposing. The upper body recedes dramatically upward into the sky or background. Head placed near top third of frame, slightly off-center.

Pose:

Grounded stance. Feet planted firmly. One foot slightly closer to the lens to enhance scale distortion. Upper body calm and controlled. No exaggerated motion. Dominance comes from stillness.

Camera Behavior:

Very close distance to foreground element. Slight horizon tilt for tension. Foreground sharp, upper portion slightly softened by depth falloff.

Lighting:

Strong upward rim light catching edges of the lower silhouette. Secondary side light sculpting volume. Subtle atmospheric haze above to create depth separation. Shadows fall downward toward the lens.

Background:

Sky-like gradient, large empty space above. Light beams or volumetric haze rising upward. No flat studio look.

Surface Detail:

Foreground textures extremely crisp. Material identity preserved exactly. No added wear unless already present.

Emotion:

Monumental. Immovable. The viewer feels physically below the character. Power through perspective alone.

Negative prompt:

No anatomy distortion, no stretched limbs beyond lens effect, no added accessories, no centered symmetry, no neutral studio lighting.
ar 1:1

Billie Eilish Cartoon Version 3D Render Prompt

game-asset

A highly structured prompt for Gemini Nano Banana Pro to generate a semi-realistic 3D character render of Billie Eilish in a cartoon version. It specifies detailed facial features, dark green hair in pigtails, specific jewelry, oversized green clothing, and technical details like 8K resolution and cinematic lighting.

展开提示词全文
{
"image_type": "3D-rendered digital portrait",
"resolution_target": "8K",
"aspect_ratio": "3:4",
"camera_framing": {
"shot_type": "medium close-up",
"orientation": "vertical",
"subject_position": "subject occupies center-left frame",
"crop": "head to upper torso visible"
},
"camera_angle": {
"head_pose": "three-quarter profile",
"face_direction": "turned to the viewer’s right",
"eye_direction": "looking slightly toward the camera"
},
"subject": {
"gender_presentation": "female",
"age_appearance": "young adult",
"skin_tone": "light skin tone",
"facial_features": {
"face_shape": "slightly elongated with soft contours",
"forehead": "smooth, broad",
"eyes": {
"color": "green",
"shape": "large, rounded",
"eyelashes": "visible, natural thickness"
},
"eyebrows": "dark, thick, naturally arched",
"nose": {
"shape": "small to medium",
"piercing": "silver hoop nose ring on left nostril"
},
"mouth": {
"expression": "subtle smirk",
"lips": "medium fullness, natural pink tone"
},
"freckles": "visible freckles across nose and cheeks",
"marks": "small scratch-like mark on right cheek"
}
},
"hair": {
"color": "{argument name="hair color" default="dark green"}",
"style": "two high pigtails",
"parting": "center part",
"texture": "smooth with loose flyaway strands",
"hair_ties": "green bands",
"loose_strands": "multiple thin strands framing the face"
},
"ears_and_jewelry": {
"ears_visible": true,
"ear_piercings": "small gold hoop earring visible on right ear",
"neckwear": [
{
"type": "choker",
"material": "black leather",
"details": "silver studs"
}
],
"necklaces": [
{
"material": "gold",
"style": "thin chain",
"pendant": "small circular ring"
},
{
"material": "gold",
"style": "thicker chain",
"pendant": "padlock-shaped charm"
}
]
},
"clothing": {
"top": {
"type": "oversized t-shirt",
"color": "green",
"fabric_appearance": "soft cotton",
"graphics": "black abstract symbol on chest",
"fit": "loose"
}
},
"pose_and_body": {
"shoulders": "slightly angled",
"neck": "elongated, relaxed",
"posture": "upright with casual tilt"
},
"lighting": {
"type": "soft studio lighting",
"direction": "front-left",
"shadows": "gentle shadows defining facial contours",
"contrast": "moderate"
},
"background": {
"color": "neutral light gray",
"texture": "smooth gradient",
"depth": "shallow depth of field",
"distractions": "none"
},
"render_style": {
"style": "semi-realistic 3D character render",
"skin_detail": "highly detailed skin texture",
"material_quality": "realistic fabric and metal shaders",
"edges": "clean, polished",
"overall_finish": "AI-generated, ultra-clean look"
},
"color_palette": {
"dominant_colors": ["green", "gray"],
"accent_colors": ["gold", "black", "silver"]
},
"quality_tags": [
"ultra-detailed",
"8K",
"high sharpness",
"cinematic lighting",
"professional 3D render"
]
}

Photorealistic Batgirl Cosplay Portrait Prompt (Exaggerated Anatomy)

game-asset

A highly explicit JSON prompt for generating a photorealistic portrait of a woman (Megan Fox resemblance) in a Batgirl-inspired bodysuit with exaggerated hourglass proportions (large bust, wide hips). It specifies the costume details (black/purple/yellow, halter-neck), pose (weight shifted, hand on granite countertop), and environment (indoor bathroom with comic poster), using negative prompts to ensure unstyled, physical proportions.

展开提示词全文
{
  "subject": {
    "desc": "Megan Fox resemblance, fair skin, long straight dark brown/black hair over shoulders/back, symmetrical face with full lips, defined brows, green-hazel eyes, tiny Batgirl mask covering eyes",
    "apparel": {
      "item": "Batgirl inspired bodysuit",
      "style": "Halter-neck sleeveless, high-cut legs",
      "color": "Black base with glossy purple accents, yellow trim on legs, large yellow bat emblem on chest",
      "details": "Tight stretch fabric emphasizing curves, black/purple stockings"
    },
    "anatomy": {
      "body_type": "Exaggerated voluptuous hourglass",
      "bust": "Large full voluminous with strong forward projection, tightly contained",
      "hips": "Very wide hips, thick thighs, narrow waist",
      "skin": "Smooth fair with natural tone variation"
    }
  },
  "pose": {
    "posture": "Standing upright, weight shifted to right hip (viewer's left) for pelvic tilt, torso slightly angled toward camera",
    "limbs": {
      "right_arm": "Extended down, hand resting on granite countertop",
      "left_arm": "Hanging naturally by side, partially obscured by hip",
      "head": "Slight tilt, facing forward"
    },
    "gaze": "Direct eye contact, soft neutral confident sultry expression"
  },
  "environment": {
    "location": "Indoor bathroom",
    "elements": [
      "Speckled green/black granite countertop (left foreground)",
      "Wooden door with tall frosted glass panel (right background)",
      "Towel bar (background left)",
      "Large Wonder Woman comic poster on black walls, small neon 'KeorUnreal' sign"
    ]
  },
  "camera": {
    "shot": "Medium (thighs up), vertical 3:4",
    "angle": "Eye-level slightly frontal",
    "lens": "50-85mm portrait",
    "dof": "Moderate, sharp subject, slightly soft distinct background",
    "framing": "Centered subject"
  },
  "lighting": "Soft diffuse indoor artificial, overhead/slightly frontal, even illumination, soft shadows under bust/hips, highlights on body contours and fabric sheen",
  "mood": "Confident calm sultry, casual intimate domestic atmosphere",
  "style": "Photorealistic smartphone photography, high-fidelity detailed fabric/skin textures, natural color grading, unstyled physical proportions",
  "colors": "Black suit with vibrant yellow/purple accents dominant, green/black granite, warm wood, neutral walls, natural contrast/saturation",
  "quality": "High-def sharp focus on eyes/face/suit pattern, low noise, minimal grain",
  "negative_prompt": "small/reduced/flattened bust, lifted breasts, slimmed torso/hips, proportion averaging, plastic/airbrushed skin, beautification, stylization, depth flattening, wide distortion, mirror selfies, phone-in-hand, reflections"
}

提示词模型无关,可直接用于 Midjourney / DALL·E / Flux / Stable Diffusion 等任意文生图模型。

信息边界说明

本页信息摘自官方文档,仅作技术信息呈现,不代表本站的判断或推荐。价格、配额、可用性以官方最新说明为准。你离开本站后的任何行为(申请 Key、付费、部署、商用),本站不做任何承诺与担保。

同类推荐

视频短视频图片音乐音频配音口播数字人全能创作

官方标注收费

通过同一个AI创作工具完成AI图片、视频、音乐与语音创作,查找公开社交数据,编辑视觉成果,并集中查看和管理生成资产。Beatra 全能创作支持AI图片生成(文生图)、AI视频生成(图生视频)和AI视频编辑,通过AI音乐生成把想法做成歌曲或纯音乐,也可用AI语音生成(文字转语音)、多语言旁白和AI语音克隆完成跨媒介内容;还能查找抖音数据、小红书数据,以及 YouTube、TikTok、Instagram 与 X 上的公开帖子、笔记、评论、账号和趋势,把社交媒体数据用于竞品调研和创作。它是一款面向AI内容创作与多模态内容创作的AI创作套件,用起来就是一个把图片、视频、音乐与语音收在同一处的AI创作助手,适用于社交内容、产品展示、营销素材、课程、播客、短片和跨媒介项目;还能在不同媒介间复用原始素材、已有成果和查到的公开资料,查看制作进度、返回文件、实际用量和积分记录。

AI数字人口播与虚拟主播视频

官方标注收费

用一张人像和一段短脚本或语音,制作有明确讲述内容的数字人口播视频。可使用现成录音,也可以先选择音色,把脚本生成旁白,让画面中的人物随声音完成稳定、自然的口播。适合 AI 主播、虚拟主播视频、数字人讲解、让照片说话、照片说话视频、产品讲解视频、数字人培训视频、公告和社交口播,并在交付后检查人物一致性、声音清晰度、口型同步和动作稳定性。

罗斯蒙德可视化图形生成

官方未说明

【罗斯蒙德可视化图形生成】南京罗斯蒙德基于已交付项目沉淀的可视化图形生成场景:根据业务描述或系统需求,自动生成可编辑的架构图、流程图、ER 图、时序图、状态图与 C4 图。采用「统一图模型」一次建模、多格式输出:结构化文本(Mermaid / PlantUML,可嵌入 Markdown、进 Git 版本管理)、可编辑图形文件(drawio / Excalidraw,可拖拽修改节点与样式)、静态图(SVG 自渲染,PNG 给转换命令),并输出图形说明(图的含义 + 关键节点解释)。内置样式系统统一配色与命名,内置布局引擎自动排布,校验脚本拦截孤立节点、悬挂连线与语义缺漏。触发词:画架构图、画流程图、画 ER 图、画时序图、画状态图、C4 图、系统架构图、部署架构图、数据流程图、实体关系图、生成 drawio、生成 Excalidraw、导出 mermaid、导出 PlantUML、导出 SVG、导出 PNG、可编辑的图、图转可编辑、架构图可编辑、流程图二次修改、画图、出图、画个图。

架构图一键生成

官方标注收费需申请 Key

使用 ContextWeave 生成或修改架构图、流程图、思维导图和复杂信息图。适用于需要从代码、文件或自然语言中提取结构与关系并实际产出 CW、SVG、HTML、可编辑 PPTX 或原生 VSDX 的请求;未特别说明时,PPTX 与 Visio 使用原生形状和连接器导出。不适用于统计图表、手绘插画或像素级排版。

AI Image Prompts · 绘画提示词库

官方标注免费

Recommend curated prompts from a 10,000+ real-world image generation prompt library. Works with ANY AI image model — Nano Banana Pro, Nano Banana 2, Seedream 5.0, GPT Image 1.5, Midjourney, DALL-E 3, Flux, Stable Diffusion, and more. Use this skill when users want to: - Find proven image generation prompts (any model) - Get prompt inspiration for portraits, products, social media, posters, etc. - Create illustrations for articles, videos, podcasts, or marketing content - Browse categorized prompt templates with sample images - Translate and understand prompt techniques

架构图一键生成

官方标注收费需申请 Key

使用 ContextWeave 生成或修改架构图、流程图、思维导图和复杂信息图。适用于需要从代码、文件或自然语言中提取结构与关系并实际产出 CW、SVG、HTML、可编辑 PPTX 或原生 VSDX 的请求;未特别说明时,PPTX 与 Visio 使用原生形状和连接器导出。不适用于统计图表、手绘插画或像素级排版。

什么是技能

技能(Skill)是封装好的单项能力,比如文字识别、文档摘要、关键词提取。开发者把调用逻辑和提示词打包,使用者像用插件一样加载即可,无需了解底层模型。启枢技能 收录大量技能,可直接试跑或编入方案。