Token导航 LogoToken导航

Agent Skill · 研究检索

create-html-carousel(create HTML carousel 搜索)

create-html-carousel 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。

create-html-carousel 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,权限为执行命令;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 52,870。

总安装 544周安装 22GitHub Stars 607下载 171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:create-html-carousel(create HTML carousel 搜索)
来源仓库:https://github.com/athina-ai/goose-skills
仓库路径:skills/create-html-carousel
安装命令:
npx skills add https://github.com/athina-ai/goose-skills --skill create-html-carousel
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/athina-ai/goose-skills --skill create-html-carousel

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、检索和筛选相关信息,适合根据关键词或任务场景快速定位候选结果。适用于 Codex、Claude、Cursor、Gemini CLI 等宿主环境,支持可视化内容生成与导出。通过命令行安装,需结合来源仓库文档核验具体用法,确保输出 PNG 格式兼容 LinkedIn。

安装前建议确认权限范围,避免触发未授权的联网或文件操作。create-html-carousel 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

来源仓库内路径:skills/create-html-carousel

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/athina-ai/goose-skills --skill create-html-car…;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

    它需要执行命令,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

create-html-carousel 怎么安装?

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

create-html-carousel 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。

create-html-carousel 需要什么权限?

权限级别是执行命令。安装流程涉及命令执行,可能通过 npx skills add https://github.com/athina-ai/goose-skills --skill create-html-carousel 联网下载 Skill 或依赖。用户安装前应确认命令来源、仓库内容和执行环境。

create-html-carousel 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。

create-html-carousel 用的人多吗?

公开数据是累计安装 544 次,近一周安装 22 次,下载 171 次,GitHub 607 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。

create-html-carousel 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

  • 需要执行命令,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

  • 不能接受 Skill 在你机器上执行命令的人。
  • 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
create-html-carousel

create HTML carousel 搜索 · 本页

544607执行命令未明确通过

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

Deprecated: This skill is superseded by goose-graphics. See skills/composites/goose-graphics/ (install with npx goose-skills install goose-graphics). The carousel format is one of seven formats in the newer skill and supports 36 style presets plus image sourcing and PNG export. This skill is retained for one release cycle before removal.

LinkedIn Carousel Creator

Create stunning LinkedIn carousel posts as PNG images. This skill generates styled HTML slides optimized for square format (1080×1080px), then automatically screenshots each slide for direct upload to LinkedIn.

Core Philosophy

  1. LinkedIn-First Design — Square format (1080×1080px), optimized for mobile feed viewing
  2. Informational Content — Tips, workflows, lists, frameworks (not presentations)
  3. Consistent Styling — Reuse proven design systems from frontend-slides
  4. Automated Export — Generate HTML → Screenshot → PNG files ready for LinkedIn
  5. Viewport Perfect — Every slide must fit exactly in 1080×1080px without scrolling

LinkedIn Carousel Specs

Format: Square (1080×1080px)

  • Aspect ratio: 1:1
  • File format: PNG (recommended) or JPG
  • File size: Under 10MB per image
  • Max slides: 10 images per carousel
  • Ideal slide count: 5-8 slides (best engagement)

Content Structure:

  1. Cover slide — Hook + title + your brand
  2. Content slides — One key point per slide (3-6 slides)
  3. Closing slide — CTA / summary / follow prompt

When to Use This Skill

Use for LinkedIn carousel posts like:

  • "5 AI GTM workflows you should be using"
  • "How to build X: A step-by-step guide"
  • "7 mistakes founders make with Y"
  • "The complete framework for Z"
  • "Before & After: How we 10x'd our metrics"

NOT for:

  • Long-form presentations (use frontend-slides)
  • Video content
  • Single-image posts

Workflow Overview

1. Content Input → User provides topic/outline
2. Style Selection → Choose visual style (or preview options)
3. HTML Generation → Create 1080×1080px HTML slides
4. Screenshot → Auto-capture each slide as PNG
5. Delivery → Folder of PNG files ready for LinkedIn upload

Phase 1: Content Discovery

Step 1.1: Get Topic & Structure

Ask the user:

Question 1: What's the topic?

  • Header: "Topic"
  • Question: "What's the main topic of this carousel?"
  • (Free text input)

Question 2: Content Type

  • Header: "Format"
  • Question: "What type of post is this?"
  • Options:

- "Numbered list" — "5 ways to...", "7 mistakes...", "3 steps to..." - "How-to guide" — Step-by-step tutorial or process - "Framework" — Concept explanation with structure - "Before/After" — Transformation or case study - "Insights/Tips" — Collection of advice or learnings

Question 3: Slide Count

  • Header: "Length"
  • Question: "How many slides?"
  • Options:

- "Short (5-6)" — Quick, punchy (best for mobile scrolling) - "Medium (7-8)" — Standard carousel length - "Long (9-10)" — Maximum LinkedIn allows

Question 4: Branding Handle

  • Header: "Brand"
  • Question: "What handle or name should appear on each slide?"
  • (Free text input — e.g., "@yourhandle", "Acme Inc", or leave blank for none)

Question 5: Content Ready?

  • Header: "Content"
  • Question: "Do you have the content written?"
  • Options:

- "Yes, I have all content" — Paste it in - "I have bullet points" — Need light formatting - "Just the topic" — Need help outlining

If user has content, ask them to share it.

Content Density Rules for LinkedIn

Each slide should be scannable in 2-3 seconds on mobile:

Slide TypeMax Content
CoverTitle (1 line) + subtitle (1 line) + branding
List itemNumber/icon + heading (2 lines max) + body (3 lines max)
FrameworkDiagram/visual + 2-4 labels
Quote/Stat1 large stat or quote + context
CTA1 action + visual element

If content exceeds limits: Break into multiple slides or simplify.


Phase 2: Style Selection

Users can choose styles two ways:

Option A: Direct Selection (Faster)

Show preset picker:

Question: Pick a Style

  • Header: "Style"
  • Question: "Which visual style works best for your content?"
  • Options:

- "Bold Signal" — High-contrast card on dark, confident - "Dark Botanical" — Elegant dark with soft abstract shapes - "Notebook Tabs" — Editorial cream paper with colorful tabs - "Pastel Geometry" — Friendly pastels with decorative pills - "Neon Cyber" — Futuristic tech aesthetic - "Split Pastel" — Playful two-tone split design

(See STYLE_PRESETS.md for full details on each style)

Option B: Guided Discovery

If user isn't sure, ask:

Question: Audience & Tone

  • Header: "Vibe"
  • Question: "Who's your audience and what tone?"
  • Options:

- "Professional/Corporate" → Recommend: Bold Signal, Dark Botanical - "Creative/Playful" → Recommend: Split Pastel, Pastel Geometry - "Technical/Dev-focused" → Recommend: Neon Cyber, Terminal Green - "Elegant/Premium" → Recommend: Dark Botanical, Paper & Ink

Then generate 2-3 preview slides and let user pick.


Phase 3: Generate HTML Carousel

File Structure

All carousel files (HTML source and PNG exports) are saved to the shared assets directory.

[carousel-name]/
├── index.html              # Full carousel (all slides)
├── slides/
│   ├── slide-01.html       # Individual slide pages
│   ├── slide-02.html
│   └── ...
└── exports/
    ├── slide-01.png        # Screenshots (generated in Phase 4)
    ├── slide-02.png
    └── ...

HTML Architecture for 1080×1080px

CRITICAL: LinkedIn carousel slides are SQUARE (1:1 ratio), not widescreen.

<!DOCTYPE html>
<html lang="en">
  <head>
    <meta charset="UTF-8" />
    <meta name="viewport" content="width=device-width, initial-scale=1.0" />
    <title>Slide 01</title>

    <!-- Fonts -->
    <link rel="stylesheet" href="https://api.fontshare.com/v2/css?f[]=..." />

    <style>
      /* ===========================================
           LINKEDIN CAROUSEL: SQUARE FORMAT
           Fixed 1080×1080px for screenshot
           =========================================== */
      :root {
        /* Fixed size for LinkedIn */
        --slide-width: 1080px;
        --slide-height: 1080px;

        /* Colors (from chosen preset) */
        --bg-primary: #0a0f1c;
        --text-primary: #ffffff;
        --accent: #00ffcc;

        /* Typography - scaled for square format */
        --title-size: 72px;
        --subtitle-size: 36px;
        --body-size: 28px;
        --small-size: 20px;

        /* Spacing */
        --slide-padding: 80px;
        --content-gap: 40px;

        /* Animation */
        --ease-out-expo: cubic-bezier(0.16, 1, 0.3, 1);
      }

      * {
        margin: 0;
        padding: 0;
        box-sizing: border-box;
      }

      html,
      body {
        width: var(--slide-width);
        height: var(--slide-height);
        overflow: hidden;
      }

      body {
        font-family: var(--font-body);
        background: var(--bg-primary);
        color: var(--text-primary);
        display: flex;
        flex-direction: column;
        justify-content: center;
        align-items: center;
        padding: var(--slide-padding);
      }

      /* Content container */
      .slide-content {
        width: 100%;
        max-width: 100%;
        display: flex;
        flex-direction: column;
        gap: var(--content-gap);
      }

      /* Typography hierarchy */
      h1 {
        font-size: var(--title-size);
        font-weight: 800;
        line-height: 1.1;
        margin-bottom: 20px;
      }

      h2 {
        font-size: var(--subtitle-size);
        font-weight: 700;
        line-height: 1.2;
      }

      p,
      li {
        font-size: var(--body-size);
        line-height: 1.4;
      }

      /* List styling */
      ul {
        list-style: none;
      }

      li {
        padding-left: 40px;
        position: relative;
        margin-bottom: 20px;
      }

      li::before {
        content: "→";
        position: absolute;
        left: 0;
        color: var(--accent);
        font-weight: bold;
      }

      /* Number badge (for list items) */
      .number {
        font-size: 120px;
        font-weight: 900;
        color: var(--accent);
        opacity: 0.15;
        position: absolute;
        top: -40px;
        left: -20px;
        z-index: 0;
      }

      /* Branding footer */
      .brand {
        position: absolute;
        bottom: var(--slide-padding);
        right: var(--slide-padding);
        font-size: var(--small-size);
        opacity: 0.7;
      }

      /* ===========================================
           STYLE-SPECIFIC OVERRIDES
           Inject preset styles here
           =========================================== */
      /* ... preset-specific CSS ... */
    </style>
  </head>
  <body>
    <div class="slide-content">
      <!-- Slide content goes here -->
      <h1>Your Title Here</h1>
      <p>Your content here</p>
    </div>

    <div class="brand">@yourbrand</div>
  </body>
</html>

Content Slide Templates

Cover Slide:

<div class="slide-content">
  <h1>5 AI GTM Workflows<br />You Should Be Using</h1>
  <p>Scale your outbound without scaling your team</p>
</div>
<div class="brand">@yourhandle</div>

Numbered Item (e.g., Slide 2/6):

<div class="slide-content">
  <div class="number">01</div>
  <h2>Signal-Based Outbound</h2>
  <p>
    Monitor job postings, funding announcements, and tech stack changes to find
    companies actively solving your problem.
  </p>
</div>
<div class="brand">@yourhandle • 1/5</div>

Framework Slide:

<div class="slide-content">
  <h2>The GTM Engineering Stack</h2>
  <div class="framework-grid">
    <div class="box">Research</div>
    <div class="box">Personalization</div>
    <div class="box">Outreach</div>
    <div class="box">Tracking</div>
  </div>
</div>
<div class="brand">@yourhandle • 3/5</div>

CTA Slide:

<div class="slide-content">
  <h2>Want more like this?</h2>
  <p>Follow me for more tips and workflows.</p>
  <div class="cta">Hit that follow button →</div>
</div>
<div class="brand">@yourhandle</div>

Phase 4: Screenshot Generation

After generating HTML, automatically capture screenshots.

Using Playwright (Recommended)

Create a Node.js script to screenshot each slide:

// screenshot-slides.js
const { chromium } = require("playwright");
const path = require("path");
const fs = require("fs");

async function screenshotSlides(slidesDir, outputDir) {
  const browser = await chromium.launch();
  const page = await browser.newPage();

  // Set viewport to LinkedIn carousel size
  await page.setViewportSize({ width: 1080, height: 1080 });

  // Ensure output directory exists
  if (!fs.existsSync(outputDir)) {
    fs.mkdirSync(outputDir, { recursive: true });
  }

  // Find all HTML files in slides directory
  const slideFiles = fs
    .readdirSync(slidesDir)
    .filter((f) => f.endsWith(".html"))
    .sort();

  console.log(`Found ${slideFiles.length} slides to screenshot`);

  for (const slideFile of slideFiles) {
    const slidePath = path.join(slidesDir, slideFile);
    const outputName = slideFile.replace(".html", ".png");
    const outputPath = path.join(outputDir, outputName);

    console.log(`Capturing ${slideFile}...`);

    await page.goto(`file://${path.resolve(slidePath)}`);

    // Wait for fonts and animations
    await page.waitForTimeout(500);

    // Take screenshot
    await page.screenshot({
      path: outputPath,
      type: "png",
      fullPage: false,
    });

    console.log(`✓ Saved ${outputName}`);
  }

  await browser.close();
  console.log("\n✨ All slides captured!");
}

// Usage
const carouselName = process.argv[2];
if (!carouselName) {
  console.error("Usage: node screenshot-slides.js <carousel-name>");
  process.exit(1);
}

const slidesDir = path.join(__dirname, carouselName, "slides");
const outputDir = path.join(__dirname, carouselName, "exports");

screenshotSlides(slidesDir, outputDir);

Installation

The skill directory needs these dependencies:

{
  "name": "linkedin-carousel-screenshots",
  "version": "1.0.0",
  "private": true,
  "dependencies": {
    "playwright": "^1.40.0"
  }
}

First time setup:

cd /path/to/skills/create-html-carousel
npm install

Running Screenshot Script

After generating HTML slides:

node screenshot-slides.js carousel-name

This will:

  1. Open each slide HTML in a headless browser
  2. Set viewport to 1080×1080px
  3. Wait for fonts/animations to load
  4. Capture PNG screenshot
  5. Save to [carousel-name]/exports/

Phase 5: Delivery

After screenshots are generated, present to user:

✨ Your LinkedIn carousel is ready!

📁 Location: /assets/carousel-name/

**Slides:**
- 6 HTML slides in slides/ folder
- 6 PNG images in exports/ folder (1080×1080px)

**Preview:**
Open index.html to see all slides with navigation.

**Upload to LinkedIn:**
1. Create new post on LinkedIn
2. Click "Add media"
3. Upload all PNGs from exports/ folder in order
4. Add your post copy
5. Publish!

**File sizes:**
- slide-01.png: 234 KB ✓
- slide-02.png: 198 KB ✓
- slide-03.png: 256 KB ✓
(All under 10MB limit)

Want to make any changes to the slides?

Style Adaptation for Square Format

All styles from frontend-slides work for carousels, but require these adjustments:

Typography Scaling

Square format has less horizontal space, so scale fonts:

ElementPresentation (16:9)Carousel (1:1)
Titleclamp(2rem, 6vw, 5rem)72px (fixed)
Subtitleclamp(1.25rem, 3vw, 2.5rem)36px (fixed)
Bodyclamp(0.875rem, 1.5vw, 1.125rem)28px (fixed)
Smallclamp(0.75rem, 1vw, 0.875rem)20px (fixed)

Why fixed sizes? We're targeting a single export size (1080×1080px), not responsive web viewing.

Layout Adjustments

Vertical space is precious:

  • Reduce top/bottom padding (80px instead of 4rem)
  • Tighter line-height (1.2-1.4 instead of 1.5-1.6)
  • Fewer list items per slide (max 3-4)
  • Smaller decorative elements

Mobile-first mindset:

  • Most LinkedIn users view on phones
  • Text must be readable at thumbnail size
  • High contrast is critical
  • Bold, simple layouts beat intricate designs

Content Best Practices

Hook Formula (Cover Slide)

Strong hooks for LinkedIn carousels:

  • Number + Promise: "5 workflows that 10x'd our outbound"
  • Contrarian: "Stop doing X. Do this instead."
  • Before/After: "How we went from X to Y in 30 days"
  • Question: "Why are only 3% of founders doing this?"
  • Curiosity gap: "The GTM strategy nobody talks about"

Body Slides (Items 2-9)

Each slide should:

  1. One clear point — Don't cram multiple concepts
  2. Visual hierarchy — Large number/icon + heading + body
  3. Concrete, not abstract — "Use job postings to find intent" not "Leverage signals"
  4. Scannable — 2-3 second read time max

Closing Slide

Always include a CTA:

  • "Follow for more [topic]"
  • "Repost if this helped"
  • "Comment your biggest takeaway"
  • "DM me if you want the full playbook"

Avoid:

  • "Link in comments" (often gets buried)
  • "Check out my website" (feels salesy)
  • No CTA at all (wasted opportunity)

Troubleshooting

Fonts Not Loading in Screenshots

Symptom: Screenshots show default system fonts

Solution:

  1. Use web-safe fonts (Arial, Georgia) OR
  2. Add await page.waitForLoadState('networkidle') before screenshot
  3. Increase wait timeout: await page.waitForTimeout(1000)

Screenshots Are Blurry

Symptom: Text looks fuzzy or low-res

Solution:

  1. Set device scale factor in Playwright: await page.setViewportSize({width: 1080, height: 1080, deviceScaleFactor: 2, // Retina-quality});

Content Overflows the Slide

Symptom: Text or elements cut off in screenshot

Solution:

  1. Reduce font sizes
  2. Decrease padding
  3. Split into multiple slides
  4. Simplify content (fewer bullets, shorter text)

Colors Look Different in Export

Symptom: PNG colors don't match HTML preview

Solution:

  • Ensure browser color profile matches sRGB
  • Use hex colors, avoid CSS filters that may render differently
  • Test screenshot script before generating all slides

Preset Quick Reference

PresetBest ForVibe
Bold SignalConfident, high-impactProfessional
Dark BotanicalElegant, premiumSophisticated
Notebook TabsEditorial, organizedFriendly-professional
Pastel GeometryFriendly, approachablePlayful
Neon CyberTech, innovationFuturistic
Split PastelCreative, funEnergetic

See STYLE_PRESETS.md for complete styling details.


Related Skills

  • frontend-slides — For full presentations (not carousels)
  • personalized-email — For outreach content to pair with LinkedIn posts
  • deep-web-research — For researching topics/stats for carousel content

Example Session Flow

  1. User: "Create a LinkedIn carousel about 5 AI GTM workflows"
  2. Skill asks: content type, slide count, have content ready?
  3. User provides bullet points of the 5 workflows
  4. Skill asks: style preference
  5. User picks "Bold Signal"
  6. Skill generates 7 HTML slides (cover + 5 workflows + CTA)
  7. Skill runs screenshot script automatically
  8. Skill delivers folder with HTML + PNG exports
  9. User uploads PNGs to LinkedIn and publishes

Total time: 5-10 minutes from idea to ready-to-publish carousel.

平台分布

Codex

36.31%
按下载量换算62

Claude

30.06%
按下载量换算51

Cursor

19.52%
按下载量换算33

Gemini CLI

9.37%
按下载量换算16

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

执行命令

安装流程涉及命令执行,可能通过 npx skills add https://github.com/athina-ai/goose-skills --skill create-html-carousel 联网下载 Skill 或依赖。用户安装前应确认命令来源、仓库内容和执行环境。

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