← 返回 Skills 市场
luoboask

Aura Image Gen

作者 luoboask · GitHub ↗ · v2.0.0 · MIT-0
cross-platform ✓ 安全检测通过
168
总下载
0
收藏
1
当前安装
5
版本数
在 OpenClaw 中安装
/install aura-image-gen
功能描述
Multi-platform infographic card generator for Xiaohongshu, Pinterest, Instagram, and more. Provides 8 visual styles × 8 layouts with platform-specific aspect...
使用说明 (SKILL.md)

Aura Image Generator (多平台图片生成)

Generate eye-catching social media card series. Style × Layout × Platform three-dimensional system.

Three Dimensions

Dimension Controls Options
Style Visual aesthetics mystic, cute, notion, bold, retro, ink-wash, tarot, minimal
Layout Information structure sparse, balanced, dense, list, comparison, flow, card, spread
Platform Aspect ratio + safe zones xiaohongshu, pinterest, instagram-square, instagram-portrait, youtube-thumbnail

Platform Specs

Platform Ratio Pixels Safe Zone Notes
xiaohongshu (默认) 3:4 1080×1440 避开右上角互动区、底部标题栏
pinterest 2:3 1000×1500 标题前 40 字符最重要,留足顶部空间
instagram-square 1:1 1080×1080 Feed 标准,中心聚焦
instagram-portrait 4:5 1080×1350 Feed 竖版,最大高度
instagram-story 9:16 1080×1920 Story/Reels,避开上下边缘
youtube-thumbnail 16:9 1280×720 大字 + 高对比,小图可读

Style Gallery

Style Description Best For
mystic (Default for astrology) 深色星空 + 金色点缀,神秘玄学感 星宿合盘、塔罗解读
cute 甜系粉嫩,经典小红书风 星座日常、轻松科普
notion 极简线条手绘,知识感 关系对照表、知识卡片
bold 高对比强冲击,大字报 避坑指南、排行榜
retro 复古怀旧,做旧质感 传统文化、国风内容
ink-wash 水墨国风,中式美学 星宿神兽、传统占卜
tarot 塔罗牌艺术风,新艺术运动边框 塔罗解读、牌面赏析
minimal 极简高端,大量留白 金句、单牌解读

Style: Mystic (详细定义)

canvas:
  ratio: portrait-3-4
  grid: single | dual

colors:
  primary: ["#0A0E2A", "#1A1040"]  # 深夜蓝/深紫
  accent: ["#C9A96E", "#D4AF37"]   # 古铜金/正金
  text: ["#E8D5B7", "#FFFFFF"]     # 暖白/纯白
  glow: ["#6B5CE7", "#A78BFA"]     # 紫色光晕

elements:
  background: starfield | nebula-gradient
  decorations: [constellation-lines, star-sparkles, moon-phases]
  frames: gold-border | arch-frame | none
  emphasis: glow-highlight | gold-underline

typography:
  title: serif-elegant | brush-calligraphy
  body: clean-sans | handwritten
  numbers: display-large

Style: Tarot (详细定义)

canvas:
  ratio: portrait-3-4
  grid: single

colors:
  primary: ["#1B0A2E", "#2D1B4E"]  # 深紫
  accent: ["#D4AF37", "#B8860B"]    # 金
  highlight: ["#8B0000", "#4A0080"] # 深红/紫
  text: ["#F5E6CC", "#FFFFFF"]

elements:
  background: velvet-dark | ornate-pattern
  decorations: [art-nouveau-border, celestial-symbols, card-frame]
  frames: tarot-border | ornate-gold
  emphasis: gold-foil | emboss

typography:
  title: art-nouveau | display-serif
  body: elegant-serif
  card-name: all-caps-spaced

Style: Ink-wash (详细定义)

canvas:
  ratio: portrait-3-4

colors:
  primary: ["#1A1A1A", "#4A4A4A"]  # 墨色
  accent: ["#8B0000", "#C41E3A"]    # 朱红
  background: ["#F5F0E8", "#E8DCC8"] # 宣纸色
  text: ["#2C2C2C", "#8B0000"]

elements:
  background: rice-paper-texture | ink-splash
  decorations: [brush-strokes, seal-stamp, ink-dots]
  frames: scroll-border | none
  emphasis: red-seal | brush-circle

typography:
  title: brush-calligraphy
  body: kai-style

Layout Gallery

Layout Density Best For
sparse 1-2 points 封面、金句、单牌
balanced 3-4 points 标准内容、性格解读
dense 5-8 points 知识卡片、对照表
list 4-7 items 排行榜、清单
comparison 2 sides 两人合盘、正逆位对比
flow 3-6 steps 流程、时间线
card Single focus 单张塔罗牌/星宿卡
spread Multi-position 牌阵展示

Language Support

Language When Notes
中文 小红书内容 默认,所有文字元素用中文
English Pinterest / international 所有文字元素用英文
双语 用户指定 同一张图可以中英双语(标题中文 + 副标题英文)

Prompt language rule:

  • AI 生图的 prompt 始终用英文(模型理解最好)
  • 图片内的展示文字跟随目标平台语言
  • Prompt 中用 Text on image in Chinese: "xxx"Text on image in English: "xxx" 明确指定

Presets (风格 + 布局 + 平台组合)

星宿内容预设

Preset Style + Layout + Platform Use Case
xingxiu-profile mystic + balanced + xiaohongshu 单宿性格解读
xingxiu-compatibility mystic + comparison + xiaohongshu 两人合盘
xingxiu-ranking bold + list + xiaohongshu 关系排行榜
xingxiu-lookup notion + dense + xiaohongshu 查宿对照表
xingxiu-beast ink-wash + card + xiaohongshu 星宿神兽卡

塔罗内容预设

Preset Style + Layout + Platform Use Case
tarot-single tarot + card + xiaohongshu 单牌解读
tarot-spread tarot + spread + xiaohongshu 牌阵解读
tarot-daily mystic + sparse + xiaohongshu 每日一牌
tarot-guide notion + list + xiaohongshu 新手教程
tarot-compare tarot + comparison + xiaohongshu 正位 vs 逆位

Pinterest 预设

Preset Style + Layout + Platform Use Case
pinterest-astrology mystic + balanced + pinterest 星座科普
pinterest-tarot tarot + card + pinterest 塔罗解读
pinterest-quote minimal + sparse + pinterest 金句灵感
pinterest-tutorial notion + flow + pinterest 教程步骤

通用预设

Preset Style + Layout + Platform Use Case
knowledge-card notion + dense + xiaohongshu 干货知识卡
warning bold + list + xiaohongshu 避坑指南
quote minimal + sparse + xiaohongshu 金句封面

Workflow

Step 1: Analyze Content

  • 识别内容类型(星宿/塔罗/知识/排行)
  • 识别目标平台(默认 xiaohongshu)
  • 推荐 preset 或 style + layout + platform 组合
  • 确定图片数量(封面 + 内容 + 结尾)

Step 2: Confirm

  • 展示推荐方案
  • 用户确认或调整

Step 3: Generate Outline

每张图的:

  • 位置(封面/内容/结尾)
  • 文字内容
  • 视觉概念
  • 使用的 layout + platform ratio

Step 4: Generate Images

  • 为每张图组装 prompt(style + layout + platform + content)
  • 参考 references/workflows/prompt-assembly.md 拼装完整 prompt
  • 使用任意 AI 生图工具执行(Midjourney/SD/DALL-E/通义万象等)
  • 第一张作为视觉锚点,后续用第一张做参考保持风格一致
  • 每张生成后报告进度

Step 5: Output

Series Complete!
Topic: [topic]
Platform: [platform] · Ratio: [ratio]
Style: [style] · Layout: [layout]
Images: N total
Files: 01-cover.png, 02-content.png, ...

Prompt Assembly

每张图的 prompt 结构:

Create a social media infographic card:

## Specs
- Platform: {platform}
- Aspect Ratio: {ratio} ({pixels})
- Style: {from style definition}

## Visual Rules
{from style yaml: colors, elements, typography}

## Layout
{from layout definition: density, structure, whitespace}

## Platform Safe Zones
{from platform specs: avoid UI overlays, title bars}

## Content
Position: {cover/content/ending}
Text: {actual content}
Visual Concept: {description}

## Language
{Match content language}

Visual Authenticity Tips

  • Add slight asymmetry to compositions
  • Offset text 2-3px from perfect center
  • Include subtle texture noise in backgrounds
  • Vary line thickness naturally
  • Allow soft edges on color blocks

References

Presets (style definitions — load when generating):

  • references/presets/mystic.md — 玄学神秘风
  • references/presets/tarot.md — 塔罗艺术风
  • references/presets/ink-wash.md — 水墨国风
  • references/presets/cute.md — 甜系少女风
  • references/presets/notion.md — 极简知识风
  • references/presets/bold.md — 高冲击力风
  • references/presets/retro.md — 复古怀旧风
  • references/presets/minimal.md — 极简高端风

Elements (canvas specs):

  • references/elements/canvas.md — 画布尺寸/安全区/布局网格
  • references/elements/platforms.md — 多平台规格详解

Workflows (process guides):

  • references/workflows/prompt-assembly.md — Prompt 拼装指南
  • references/workflows/multi-platform.md — 多平台适配指南
安全使用建议
This skill is an offline set of templates and runtime instructions for assembling prompts and layouts; it does not itself call external services or require secrets. Before installing, consider: (1) the agent will still need access to an image-generation backend (your chosen API or local model)—confirm which service will receive the prompts and review that service's privacy and usage terms; (2) test the skill with non-sensitive sample content so you can see what external API calls your agent makes in practice; (3) check copyright and asset/font usage if you plan to include third-party imagery or brand logos; (4) if you prefer tighter control, keep autonomous invocation disabled for this skill or audit logs of calls the agent makes to your image API. Overall the skill is coherent and proportionate to its stated purpose.
功能分析
Type: OpenClaw Skill Name: aura-image-gen Version: 2.0.0 The aura-image-gen skill bundle is a comprehensive framework for an AI agent to generate structured prompts for social media infographics. It contains detailed design specifications, platform-specific safe zones, and style presets (e.g., mystic, tarot, ink-wash) across files like SKILL.md and various reference documents. No evidence of data exfiltration, unauthorized network access, or malicious prompt injection was found; the bundle functions entirely as a template-driven design assistant.
能力评估
Purpose & Capability
The name and description (multi-platform infographic/card generator and prompt templates) match the actual content: style/layout presets, platform aspect-ratios, and prompt assembly workflows. There are no unrelated requirements (no binaries, env vars, or external creds).
Instruction Scope
SKILL.md and the reference files describe how to analyze content, assemble prompts, adapt layouts, and batch-generate images via external image-generation calls. All instructions stay within the stated purpose and reference only the included presets and platform guidelines; they do not instruct reading unrelated system files, environment variables, or sending data to unknown endpoints.
Install Mechanism
No install spec or code files are present (instruction-only). Nothing will be written to disk or downloaded by the skill itself, which minimizes installation risk.
Credentials
The skill declares no required environment variables, credentials, or config paths. The guidance relies on the agent's existing image-generation tool (implicitly), which is proportionate to the skill's purpose.
Persistence & Privilege
The skill is not always-enabled and does not request elevated persistence or attempts to modify other skills. Autonomous invocation is allowed by platform default but is not combined with other red flags here.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install aura-image-gen
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /aura-image-gen 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v2.0.0
Aura Image Gen 2.0.0 introduces a major upgrade with a flexible, multi-platform system and expanded style/layout options: - Added support for 8 visual styles × 8 layouts × 6+ platforms (Xiaohongshu, Pinterest, Instagram, YouTube). - Provides detailed style definitions and layout structures for rich customization. - Features a workflow system for analyzing content, selecting presets, and assembling prompts per platform specs. - New structure includes platform-specific ratios, style yaml blocks, and visual authenticity tips for professional results. - Extensive reference and preset documentation now included for styles, elements, and workflows. - Greatly enhances use cases beyond Xiaohongshu to support a wider range of social media visual needs.
v1.3.0
Repositioned as prompt engineering guide instead of API wrapper. Removed provider-specific config references. Added clear explanation that this provides prompt templates, not API integrations.
v1.2.0
Cleaned up provider config references, removed raw env var names to avoid suspicious pattern detection
v1.1.0
Added language handling rules for multi-platform image generation.
v1.0.0
Aura Image Gen 1.0.0 – Initial Release - Multi-provider AI image generation tool for social media content. - Automatically selects the best available provider based on API keys. - Presets and prompt templates optimized for Xiaohongshu, Pinterest, and Instagram formats. - Supports text-to-image, reference images for style consistency, and batch image generation. - CLI usage examples with aspect ratio, quality, provider, and social media presets provided. - Serves as backend for other content and image generation tools in the ecosystem.
元数据
Slug aura-image-gen
版本 2.0.0
许可证 MIT-0
累计安装 1
当前安装数 1
历史版本数 5
常见问题

Aura Image Gen 是什么?

Multi-platform infographic card generator for Xiaohongshu, Pinterest, Instagram, and more. Provides 8 visual styles × 8 layouts with platform-specific aspect... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 168 次。

如何安装 Aura Image Gen?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install aura-image-gen」即可一键安装,无需额外配置。

Aura Image Gen 是免费的吗?

是的,Aura Image Gen 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Aura Image Gen 支持哪些平台?

Aura Image Gen 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Aura Image Gen?

由 luoboask(@luoboask)开发并维护,当前版本 v2.0.0。

💬 留言讨论