Nano Banana Prompt Skill

v1.0.0

Converts Chinese image descriptions into structured English JSON prompts for Nano Banana's Gemini model, supporting products, avatars, logos, and social medi...

0· 18·0 current·0 all-time
byxiaoke321@xk103295870-alt
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
The name and description claim to produce structured Nano Banana (Gemini) prompts from Chinese input; the SKILL.md contains exactly that logic and example outputs. No unrelated credentials, binaries, or config paths are requested.
Instruction Scope
SKILL.md describes extracting fields from user text and constructing a JSON payload for image generation. It does not instruct the agent to read local files, environment variables, or send data to unexpected endpoints. References to Vertex AI / Google AI Studio are informational only.
Install Mechanism
No install spec and no code files are included (instruction-only), so nothing is written to disk or downloaded during install.
Credentials
The skill declares no required environment variables or credentials. Although it references Vertex AI / Google AI Studio APIs, it does not request API keys or other secrets — this is proportionate for a prompt-generation helper.
Persistence & Privilege
always is false and the skill does not request persistent or elevated privileges. It does not modify other skills or system configuration.
Assessment
This skill is instruction-only and appears coherent for converting Chinese descriptions into structured prompts for a Gemini image model. Before installing or using it, be aware: (1) it only generates JSON prompts — actually calling Vertex AI / Nano Banana would require you to supply API credentials separately; (2) if you later wire the generated JSON into an automated system that calls external APIs, review where those calls are sent and which credentials are used; and (3) treat any generated prompts containing sensitive personal data cautiously before sending them to external image-generation services.

Like a lobster shell, security has layers — review code before you run it.

latestvk97etqzp7ewqydhvpr3vd2s72x845q65

License

MIT-0
Free to use, modify, and redistribute. No attribution required.

SKILL.md

Nano Banana Prompt Generator Skill

版本:1.0.0 用途:用户中文描述 → Nano Banana 图片生成的结构化 JSON 参数


触发词

生成图片提示词、生成 Nano Banana 参数、制作产品图、设计头像、设计 LOGO、生成海报、生成配图

只要用户描述了想要生成的图片内容,即触发此 Skill。


模型背景

Nano Banana 是 Google 的 Gemini 图片生成模型(gemini-3.1-flash-image),具有:

  • 多模态理解(中英文提示词)
  • 精确的文字渲染(支持中英双语)
  • 快速生成(Flash 速度)
  • 适合产品图、头像、LOGO、海报、社交媒体配图

API 来源:


工作流程

Step 1:理解用户意图

从用户描述中提取:

字段说明来源
use_case用例类型从描述判断
subject主体(人物/产品/文字/场景)用户描述
context环境/背景用户描述
style视觉风格用户指定或推断
colors主色调用户指定或推断
text_content文字内容(如有)用户描述
aspect_ratio画布比例根据用例推断
language文字语言中文/英文/双语

Step 2:分类用例

自动归类为以下四种之一:

用例说明默认比例
product_display产品展示海报1:1 或 4:3
social_media社交媒体配图16:9 或 9:16
logo_designLOGO 设计1:1
avatar头像/肖像1:1

Step 3:生成结构化 JSON

输出标准 Nano Banana API 参数:

{
  "model": "imagegeneration@latest",
  "prompt": {
    "text": "...",
    "style": "...",
    "composition": "...",
    "lighting": "...",
    "mood": "..."
  },
  "parameters": {
    "numberOfImages": 1,
    "aspectRatio": "1:1",
    "outputFormat": "png",
    "safetySetting": "block_some",
    "personGeneration": "allow_adult"
  },
  "use_case": "...",
  "language_prompt": "...",
  "notes": "..."
}

Step 4:输出给用户

语言要求:

  • prompt.text → 英文(Nano Banana 英文效果最佳)
  • language_prompt → 中文原意对照(便于用户理解)
  • 如果包含文字,单独标注 text_content 字段

用例详细规范

1. 产品展示海报(product_display)

触发词: 产品图、产品海报、商品展示、电商主图

Prompt 要素:

[产品主体描述], on [背景/场景], [光照描述], [风格描述], product photography, high resolution, clean background

默认参数:

{
  "use_case": "product_display",
  "parameters": {
    "aspectRatio": "1:1",
    "numberOfImages": 1
  }
}

2. 社交媒体配图(social_media)

触发词: 微博配图、小红书、Instagram、朋友圈、公众号封面

Prompt 要素:

[场景/主题描述], [情感氛围], [构图], [风格滤镜], social media photography, mobile wallpaper

比例参考:

  • 微博/公众号封面:16:9
  • 小红书/Instagram:1:1 或 4:5
  • 朋友圈:9:16 或 1:1

3. LOGO 设计(logo_design)

触发词: LOGO、标志、品牌标识、图标

Prompt 要素:

[品牌名/概念描述], [行业属性], [风格:minimal/retro/tech/futuristic], [颜色], logo design, vector style, clean, professional

文字处理:

  • 如果 LOGO 包含文字,text_content 字段单独输出
  • Nano Banana 支持中英文字渲染,可直接写入 prompt

4. 头像/肖像(avatar)

触发词: 头像、个人照片、肖像、证件照

Prompt 要素:

[人物描述:性别/年龄/发型/表情/姿态], [背景:纯色/渐变/场景], [风格:realistic/cartoon/illustration], [光照], portrait photography, [用途]

风格参考表

风格英文关键词
写实摄影realistic photography, high fidelity
电影感cinematic, film still
扁平插画flat illustration, vector art
赛博朋克cyberpunk, neon lights
复古retro vintage, 80s aesthetic
极简minimalism, clean design
油画oil painting, impressionist
水彩watercolor style
动漫anime style, manga
像素pixel art, 8-bit
3D 渲染3D render, C4D
中国风Chinese ink style, traditional

输出格式模板

完整 JSON 输出示例

{
  "model": "imagegeneration@latest",
  "prompt": {
    "text": "A minimalist product photography of a ceramic coffee mug on a white marble surface, soft natural lighting from the left, clean background, high resolution, studio quality",
    "style": "product photography, minimalist, clean",
    "composition": "centered product, 2/3 negative space",
    "lighting": "soft natural light, subtle shadow",
    "mood": "calm, modern, premium"
  },
  "parameters": {
    "numberOfImages": 1,
    "aspectRatio": "1:1",
    "outputFormat": "png",
    "safetySetting": "block_some",
    "personGeneration": "allow_adult"
  },
  "use_case": "product_display",
  "text_content": null,
  "language_prompt": "极简风格产品图:白色大理石台面上的陶瓷咖啡杯,左侧柔和自然光,高清干净背景",
  "notes": "建议搭配产品名 LOGO 或文字使用 Nano Banana 的文字渲染功能"
}

注意事项

  1. 文字渲染: Nano Banana 支持中英文文字,直接在 prompt 中写入即可
  2. 比例选择: 默认 1:1,社交媒体配图根据平台选择合适比例
  3. 风格融合: 可以组合多个风格(如 "cyberpunk + minimalist")
  4. 安全设置: 默认 block_some,如需调整需明确说明
  5. 输出语言: prompt.text 始终输出英文,中文原意记录在 language_prompt 字段

参考资源

Files

2 total
Select a file
Select a file to preview.

Comments

Loading comments…