Token导航 LogoToken导航

Agent Skill · 研究检索

icon-generation(图标生成)

icon-generation 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

icon-generation 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为接触敏感数据;许可证信息还不明确,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 52,401。

总安装 559周安装 24GitHub Stars 10下载 196核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:icon-generation(图标生成)
来源仓库:https://github.com/michaelboeding/skills
仓库路径:skills/icon-generation
安装命令:
npx skills add https://github.com/michaelboeding/skills --skill icon-generation
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/michaelboeding/skills --skill icon-generation

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

用于查找、检索和筛选与图标生成相关的资源或工具链。适合在开发前期快速定位可用素材或自动化生成方案。icon-generation 属于研究检索类 Skill,可作为该场景下的辅助能力补充。需根据项目需求过滤结果,避免引入不兼容格式或授权问题。可通过 GitHub 仓库安装,集成至 Codex、Claude 等宿主环境。

使用前应确认来源仓库是否提供实际生成逻辑而非仅列表信息。

来源仓库内路径:skills/icon-generation

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/michaelboeding/skills --skill icon-generation;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要接触敏感数据,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

icon-generation 怎么安装?

在终端执行 npx skills add https://github.com/michaelboeding/skills --skill icon-generation。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

icon-generation 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

icon-generation 需要什么权限?

权限级别是接触敏感数据。该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

icon-generation 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

icon-generation 用的人多吗?

公开数据是累计安装 559 次,近一周安装 24 次,下载 196 次,GitHub 10 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

icon-generation 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 需要接触敏感数据,装到常用环境前要先评估影响面。
  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 环境里有大量密钥、凭证,不方便给第三方 Skill 读取的人。
  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
icon-generation

图标生成 · 本页

55910接触敏感数据未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Icon Generation Skill

Generate professional app icons with transparent backgrounds using AI (Google Gemini).

Output: Square PNG with transparent background - ready for app stores, favicons, or UI.

Quick Examples

User SaysWhat Happens
"Create an icon for a music app"Generates square PNG with transparent background
"Make a settings gear icon, flat style"Flat design icon with clean lines
"Generate a 3D shopping cart icon"3D-style icon with depth and shadows
"I need a favicon for my blog"Small icon optimized for web
"Create icon set: home, search, profile"Batch generates multiple icons
"Make icons matching this style" + imageUses reference image to match style
"Generate more icons like this one"Creates consistent icon set from reference

Prerequisites

  • GOOGLE_API_KEY - Required for Google Gemini image generation

- Get your key at: https://aistudio.google.com/apikey

Icon Styles

StyleDescriptionBest For
flatModern flat design, solid colors, minimal shadowsiOS, Material Design, web
3dDepth, gradients, realistic shadowsmacOS, premium apps
lineOutline only, thin strokesMinimalist UI, toolbars
glyphSolid filled shape, single colorSystem icons, tab bars
gradientSmooth color transitionsModern app icons
minimalUltra-simple, essential shapes onlyProfessional, clean UI

Workflow

Step 1: Gather Requirements (REQUIRED)

Use the AskUserQuestion tool for each question. Ask ONE question at a time.

Q1: Icon Concept

"What should this icon represent? Examples: settings gear, shopping cart, user profile, home, music note, camera"

*Wait for response.*

Q2: Style

"What style? - Flat (modern, solid colors) - 3D (depth, shadows) - Line (outline only) - Glyph (solid filled shape) - Gradient (color transitions) - Minimal (ultra-simple)"

*Wait for response.*

Q3: Size

"What size? (can generate multiple) - 1024px (App Store, high-res) - 512px (Android, web) - 256px (desktop apps) - 128px (favicons, small UI) - All common sizes"

*Wait for response.*

Q4: Reference Image (Optional but important for icon sets)

"Do you have a reference image to match the style? - Yes, I have an existing icon to match - No, generate fresh style"

*If yes, ask for the image path. Wait for response.*

Q5: Colors (Optional)

"Any color preferences? - Let AI choose based on concept - Specific colors (e.g., blue and white) - Monochrome (single color + transparency) - Match brand colors"

*Wait for response.*

Q6: Output Format

"What file format? - PNG (transparency, recommended) - JPEG (smaller file, no transparency) - WebP (modern format, good compression)"

*Wait for response.*

Q7: Background Removal Method (for PNG/WebP)

"How should we remove the background? - Built-in (fast, may have minor artifacts) - Rembg (AI-based, high quality, runs locally) - None (keep white background from generation)"

*Wait for response.*

Step 2: Craft the Prompt

Transform user request into an icon-specific prompt:

Template:

A [STYLE] style icon of [CONCEPT] on a transparent background.
Square format, centered, clean edges, suitable for app icon.
[COLOR PREFERENCES]. Simple, recognizable, professional.

Example transformations:

  • User: "music app icon"
  • Enhanced: "A flat style icon of a musical note on a transparent background. Square format, centered, clean edges, suitable for app icon. Vibrant purple and pink gradient. Simple, recognizable, professional design with modern aesthetic."
  • User: "settings gear, minimal"
  • Enhanced: "A minimal style icon of a settings gear on a transparent background. Square format, centered, clean edges. Monochrome gray. Ultra-simple design, essential shapes only, no extra details."

Step 3: Generate the Icon

Execute the script:

python3 ${CLAUDE_PLUGIN_ROOT}/skills/icon-generation/scripts/icon_generator.py \
  --prompt "your enhanced prompt" \
  --style "flat" \
  --size 1024 \
  --output "/path/to/icon.png"

With multiple sizes:

python3 ${CLAUDE_PLUGIN_ROOT}/skills/icon-generation/scripts/icon_generator.py \
  --prompt "your enhanced prompt" \
  --style "flat" \
  --size 1024 512 256 128 \
  --output "/path/to/icon"

This generates: icon_1024.png, icon_512.png, icon_256.png, icon_128.png

Batch generation (multiple icons):

python3 ${CLAUDE_PLUGIN_ROOT}/skills/icon-generation/scripts/icon_generator.py \
  --batch '["home", "search", "profile", "settings"]' \
  --style "flat" \
  --size 512 \
  --output "/path/to/icons"

This generates: home_512.png, search_512.png, profile_512.png, settings_512.png

Step 4: Deliver the Result

  1. Show the generated icon(s) to the user
  2. Provide the prompt used
  3. Offer to:

- Generate different sizes - Try a different style - Create variations - Generate as ICO (favicon format)

Script Parameters

ParameterShortDescriptionDefault
--prompt-pIcon descriptionRequired
--style-sIcon style presetflat
--size-zSize(s) in pixels512 (or auto from reference)
--format-fOutput format (png, jpeg, webp)png
--bg-removal--bgBackground removal method (builtin, rembg, none)builtin
--output-oOutput pathauto-generated
--batch-bJSON array of conceptsNone
--colors-cColor preferencesAI choice
--reference-rReference image(s) for style matchingNone

Background Removal Methods

This skill uses the background-remove skill for transparent background generation.

MethodDescriptionBest For
builtinFast white-to-transparent conversionQuick iterations, most icons
rembgAI-based removal using U2-Net modelHigh-quality results, complex backgrounds
noneKeeps the white background as-isJPEG output, manual post-processing

Requirements

See the background-remove skill for installation instructions.

# Install rembg for AI-based background removal
pip install rembg

# Or with GPU acceleration (requires CUDA)
pip install rembg[gpu]

Style Prompt Modifiers

The script automatically adds these style-specific prompt modifiers:

StyleModifier Added
flat"flat design, solid colors, no gradients, minimal shadows, clean vector style"
3d"3D style, depth, soft shadows, subtle gradients, professional rendering"
line"line art, outline only, thin consistent strokes, no fill, minimalist"
glyph"solid filled shape, single color, bold silhouette, high contrast"
gradient"smooth gradient colors, modern vibrant transitions, glossy finish"
minimal"ultra-minimal, essential shapes only, maximum simplicity, geometric"

Common Icon Sizes

SizeUse Case
1024pxiOS App Store, high-resolution displays
512pxAndroid Play Store, macOS apps
256pxWindows apps, desktop shortcuts
192pxAndroid launcher icons
180pxiOS home screen
128pxFavicons (high-res), small UI
64pxToolbar icons
32pxStandard favicon
16pxSmallest favicon

Error Handling

Missing API key:

Error: GOOGLE_API_KEY not set. Get your key at https://aistudio.google.com/apikey

Generation failed:

  • Retry with simpler prompt
  • Check if concept is clear and recognizable
  • Ensure style is appropriate for the concept

Non-transparent output:

  • The script explicitly requests transparent background
  • If still not transparent, try adding "isolated on transparent" to prompt
  • For complex concepts, results may vary

Reference Image Support

Use reference images to create consistent icon sets. When you provide a reference, the AI will match:

  • Visual style and design language
  • Line weight and stroke thickness
  • Color palette and shading
  • Level of detail and complexity
  • Overall proportions and spacing

When to Use Reference Images

ScenarioUse Reference?
Creating multiple icons for same appYes - ensures consistency
Matching existing brand iconsYes - maintains brand identity
Adding to existing icon setYes - blends seamlessly
Single standalone iconOptional - fresh style is fine
Exploring different stylesNo - let AI be creative

Reference Image Examples

Single icon matching existing style:

python3 ${CLAUDE_PLUGIN_ROOT}/skills/icon-generation/scripts/icon_generator.py \
  -p "settings gear" \
  -r /path/to/existing_icon.png \
  -z 512

Icon set with consistent style:

python3 ${CLAUDE_PLUGIN_ROOT}/skills/icon-generation/scripts/icon_generator.py \
  -b '["home", "search", "profile", "settings", "notifications"]' \
  -r /path/to/brand_icon.png \
  -z 512 \
  -o "app_icons"

Multiple reference images (up to 14):

python3 ${CLAUDE_PLUGIN_ROOT}/skills/icon-generation/scripts/icon_generator.py \
  -p "calendar" \
  -r /path/to/icon1.png \
  -r /path/to/icon2.png \
  -z 512

Tips for Best Results

  1. Be specific about the concept: "shopping cart" is better than "e-commerce"
  2. Match style to platform: Flat for iOS/Android, 3D for macOS
  3. Keep it simple: Icons should be recognizable at small sizes
  4. Test at small sizes: Generate 32px to ensure readability
  5. Use consistent style: For icon sets, use same style for all icons
  6. Use reference images: For icon sets, always use a reference to ensure consistency

Examples

App Icon

python3 ${CLAUDE_PLUGIN_ROOT}/skills/icon-generation/scripts/icon_generator.py \
  -p "camera with colorful lens" \
  -s gradient \
  -z 1024 512 256 \
  -o "camera_icon"

Favicon

python3 ${CLAUDE_PLUGIN_ROOT}/skills/icon-generation/scripts/icon_generator.py \
  -p "letter B for blog" \
  -s minimal \
  -z 128 32 16 \
  -o "favicon"

Icon Set

python3 ${CLAUDE_PLUGIN_ROOT}/skills/icon-generation/scripts/icon_generator.py \
  -b '["home", "search", "notifications", "profile", "settings"]' \
  -s flat \
  -z 512 \
  -o "app_icons"

High-Quality Background Removal (Rembg)

python3 ${CLAUDE_PLUGIN_ROOT}/skills/icon-generation/scripts/icon_generator.py \
  -p "detailed gear with shadows" \
  -s 3d \
  -z 512 \
  --bg rembg \
  -o "gear_icon"

WebP Format with Built-in Background Removal

python3 ${CLAUDE_PLUGIN_ROOT}/skills/icon-generation/scripts/icon_generator.py \
  -p "lightning bolt" \
  -s gradient \
  -z 512 \
  -f webp \
  -o "bolt_icon"

平台分布

Codex

34.47%
按下载量换算68

Claude

32.34%
按下载量换算63

Cursor

20.87%
按下载量换算41

Gemini CLI

9.83%
按下载量换算19

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

接触敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills