Token导航 LogoToken导航

Agent Skill · 研究检索

visual-note-card(视觉记录卡)

用于辅助界面设计、视觉规范、排版、配色、布局和交互体验优化。它适合让 Agent 根据产品场景整理页面结构、生成 UI 方案、检查视觉一致性或改进组件层级。使用时需要结合现有品牌、设计系统和用户任务,不应只堆装饰元素;涉及真实页面改动时,应通过截图或浏览器预览检查文本溢出、对齐和响应式表现。

visual-note-card 是研究检索类 Agent Skill,可用宿主为 OpenClaw,通过ClawHub获取,权限为操作浏览器;许可证为 MIT-0,第三方扫描有提醒项,按安装量在研究检索类的 73,675 个 Skill 里排第 16,838。

总安装 6,961周安装 299GitHub Stars 公开资料未说明下载 2,440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
MIT-0
版本
1.0.0

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:visual-note-card(视觉记录卡)
来源仓库:https://github.com/tom859174-sketch/visual-note-card
安装命令:
openclaw skills install visual-note-card
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 OpenClaw 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。

ClawHubOpenClaw
openclaw skills install visual-note-card

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。

生成专业的中文视觉笔记卡作为单页信息图表并导出 PNG。适用于界面设计、视觉规范和 UI 方案整理等场景。输入内容描述即可自动生成排版美观的视觉卡片。visual-note-card 属于研究检索类 Skill,可作为该场景下的辅助能力补充。需结合品牌和设计系统使用,避免堆砌装饰元素。

涉及页面改动时应通过截图检查文本溢出和对齐表现。

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 OpenClaw;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 openclaw skills install visual-note-card;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

    它需要操作浏览器,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

visual-note-card 怎么安装?

在终端执行 openclaw skills install visual-note-card。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

visual-note-card 支持哪些 AI 工具?

公开资料标注支持 OpenClaw。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。

visual-note-card 需要什么权限?

权限级别是操作浏览器。该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

visual-note-card 是免费的吗?

来源标注的许可证是 MIT-0。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

visual-note-card 用的人多吗?

公开数据是累计安装 6,961 次,近一周安装 299 次,下载 2,440 次。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。

visual-note-card 安全吗?

第三方扫描有提醒项,说明存在需要人工确认的行为。安装前请打开源码看清它读了什么、联网到哪里。

值得选的地方和要先想清楚的地方

值得选的地方

  • 专注 OpenClaw 一种宿主,触发方式和使用习惯不用来回切换。
  • 许可证明确为 MIT-0,商用和分发边界好判断。

先想清楚的地方

  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 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 OpenClaw,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

  • 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
visual-note-card

视觉记录卡 · 本页

6,961—操作浏览器MIT-0有提醒

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

name
visual-note-card
description
Generate professional Chinese visual note cards (视觉笔记卡片/信息图) as single-page HTML infographics with automatic PNG export. Use this skill whenever the user asks to create a visual note, knowledge card, infographic, one-pager summary, visual summary, 知识卡片, 视觉笔记, 信息图, 一页纸总结, or any poster-style knowledge visualization. Also trigger when the user wants to summarize an article, blog post, book chapter, or concept into a structured visual card format, or when they reference an existing visual note and ask to create one in the same style. This skill produces both a self-contained HTML file and a high-quality PNG image, ready for sharing on social media or printing.

Visual Note Card Generator

Create professional, information-dense visual note cards (视觉笔记/信息图) as self-contained HTML files with automatic PNG export via Playwright. The output is a structured, poster-style infographic with modular card layout, suitable for social media sharing or printing.

Design System

Before writing any code, read the reference template at assets/template.html for the canonical HTML/CSS structure. Then adapt the content to the user's topic.

Core Visual Identity

The design follows an editorial knowledge card aesthetic — high information density with clear visual hierarchy, inspired by premium magazine layouts and structured note-taking.

Color Palette (CSS Variables):

  • --primary: Deep teal #1a7a6d — headers, badges, accent elements
  • --primary-dark: #145f54 — bottom highlight bar
  • --primary-deep: #0d3d36 — deep accent
  • --accent: Orange #e8713a — emphasis, secondary badges, important labels
  • --bg: Warm gray #f0ebe4 — page background
  • --bg-light: #f5f1ec — light panel background
  • --bg-card: #e8e3dc — framework card background
  • --black: #1a1a1a — dark panel, primary text
  • --white: #ffffff

Users may request alternate color schemes. If so, maintain the same structural contrast ratios: one warm neutral background, one strong primary color, one accent color, and black for dark panels.

Typography:

  • English display: Playfair Display (serif, 700/900) — main title
  • Chinese body: Noto Sans SC (400/500/700/900) — all Chinese text
  • Monospace: JetBrains Mono (500/700) — labels, URLs, tags
  • Import via Google Fonts in the <style> block

Layout — Fixed Poster Dimensions:

  • Container width: 1200px
  • Padding: 36px 40px 28px
  • Background: var(--bg)

Mandatory Layout Sections (Top to Bottom)

Every visual note card MUST include these sections in order:

1. Top Bar

A single-line flex row with:

  • Left: Topic label in uppercase monospace (e.g., AI PRODUCT ARCHITECTURE · SYSTEM DESIGN)
  • Right: Source label (e.g., OPENAI ENGINEERING BLOG)
2. Main Title Area

Two-column flex layout:

  • Left: Large bilingual title — English line in Playfair Display 38px, Chinese line in Noto Sans SC 40px colored with var(--primary)
  • Right: A 2–3 line thesis statement (核心观点) summarizing the card's argument, with the key phrase in <strong>. Optionally include a source URL below.
3. Framework Row (Flexible Grid, Recommended 4 Columns)

A row of equal-width cards representing the core framework/model of the topic. The number of columns is flexible — choose based on the actual content:

  • 2 columns — binary concepts, comparisons (e.g., Before vs After, Input vs Output)
  • 3 columns — triads, triangular models (e.g., People-Process-Technology)
  • 4 columns — recommended default, most common for frameworks (e.g., E-K-C-F, M-P-D-G)
  • 5 columns — five-element models (e.g., HEART metrics, Five Forces)
  • 6 columns — extended taxonomies (use sparingly, keep descriptions very short)

Analyze the topic and choose the column count that best fits the natural structure. Don't force 4 columns if the concept has 3 or 5 natural parts.

CSS implementation: Use grid-template-columns: repeat(N, 1fr) where N is the chosen column count.

Card structure (same regardless of column count):

  • Each card has a colored square letter badge (first letter of the concept) + Chinese name
  • Below the badge: 1–2 lines of description with one <strong> keyword
  • For 5–6 columns, keep descriptions to 1 line to avoid overflow

Badge color rotation (cycles through these in order):

  1. --primary (teal)
  2. --primary (teal)
  3. --accent (orange)
  4. --primary-deep (deep teal)
  5. --accent (orange) — if 5th column exists
  6. --primary (teal) — if 6th column exists

The framework should be a memorable acronym (e.g., E-K-C-F, M-P-D-G). Invent one if the source doesn't provide it.

4. Two-Column Content Area

A grid-template-columns: 1fr 1fr layout:

Left: Dark Panel (background: var(--black), white text)

  • Section title with emoji icon (⚡, 🔥, 🛠, etc.)
  • 2 sub-blocks, each with an orange title (var(--accent)) and a bulleted list (custom ■ bullets in teal)
  • List items use <strong> for key phrases (white color)
  • A bottom "conclusion" block with a divider line and a memorable quote/insight (key phrase in #5ee6d0 mint color)

Right: Light Panel (background: var(--bg-light))

  • Section title with star icon (★)
  • 3–4 numbered insight items, each with:

- A large serif number (Playfair Display 36px, teal) - Bold title line - 1–2 lines of description with <strong> keywords

5. Bottom Highlight Bar

Full-width bar with background: var(--primary-dark):

  • Left: The framework formula (e.g., Architecture = M × P × D × G) in mint highlight color #5ee6d0
  • Right: A closing thought in lighter text
6. Footer

Flex row:

  • Left: Framework label in monospace
  • Right: Brand / framework name with a small teal dot separator

Content Strategy

When the user provides a topic (or an article URL/text):

  1. Extract or synthesize a 4-part framework — find the core structural model. If one doesn't exist, create a meaningful decomposition with a memorable acronym.
  2. Write a provocative thesis — the right-side subtitle should be a strong, opinionated claim, not a bland description.
  3. Dark panel = "Story + Transformation" — use this for narrative content: problems, transitions, role changes, paradigm shifts.
  4. Light panel = "Pitfalls or Insights" — use this for actionable numbered takeaways.
  5. Bottom formula — distill the entire card into one equation-style summary.
  6. All text is Chinese except for: the main English title line, technical terms, framework acronyms, and footer labels.

Output

Default: HTML + PNG

By default, generate both an HTML file and a PNG image:

  1. Generate the HTML — single self-contained .html file with all CSS inline. No external dependencies except Google Fonts and html2canvas CDN. Save to /mnt/user-data/outputs/.
  2. Generate the PNG — run the bundled scripts/html2png.py script to render the HTML into a high-quality PNG image.
python <skill-path>/scripts/html2png.py /mnt/user-data/outputs/card.html /mnt/user-data/outputs/card.png --scale=1.5

Present both files to the user (PNG first, then HTML). The PNG is the primary deliverable for sharing; the HTML enables further browser-based export or editing.

If the user explicitly asks for "只要 HTML" or "HTML only", skip the PNG step.

PNG Generation Script: scripts/html2png.py

A Playwright-based headless renderer. It opens the HTML in Chromium, waits for Google Fonts to load, hides the FAB button, then screenshots the .poster element.

Usage:

python html2png.py <input.html> [output.png] [--scale=N]

Scale options:

  • --scale=1 — standard (1200px wide), smallest file
  • --scale=1.5 — default, recommended for social media (1800px wide)
  • --scale=2 — print quality (2400px wide)

Dependencies: playwright (pip install playwright && playwright install chromium). Pre-installed on Claude's compute environment.

Download Button (Mandatory in HTML)

Every generated HTML card MUST include a floating action button (FAB) in the bottom-right corner with a dropdown menu for export options:

  • 标准 PNG — 1× scale, quick sharing
  • 高清 PNG — 1.5× scale, social media recommended
  • 超清 PNG — 2× scale, for printing
  • JPEG — 1.5× scale, smaller file size

Implementation:

  1. Add <script src="https://cdnjs.cloudflare.com/ajax/libs/html2canvas/1.4.1/html2canvas.min.js"></script> before </body>
  2. The FAB (div.fab-wrap) is placed OUTSIDE .poster so it won't appear in the exported image
  3. Clicking the FAB toggles a dropdown menu upward; clicking outside closes it
  4. During capture, the FAB is hidden and a toast spinner ("正在导出…") is shown
  5. The FAB icon toggles between a download arrow and a close ×
  6. The exported filename is derived from document.title + scale suffix

See assets/template.html for the complete FAB HTML, CSS, and JS.


## Example Prompts → Expected Behavior

| User Says | Action |
|-----------|--------|
| "帮我做一张关于 RAG 架构的视觉笔记" | Generate HTML + PNG about RAG architecture |
| "把这篇文章做成信息图" + article text | Extract key ideas, synthesize framework, generate HTML + PNG |
| "生成一张同风格的卡片,主题是微服务" | Generate HTML + PNG about microservices |
| "Create a visual note about product-market fit" | Generate bilingual HTML + PNG about PMF |
| "只要 HTML,不要图片" | Generate HTML only, skip PNG |
| "生成一张 2x 高清的 PNG" | Generate HTML + PNG with `--scale=2` |

平台分布

OpenClaw

88.98%
按下载量换算2,171

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

可疑

权限和风险

操作浏览器

该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

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。来源安全扫描存在提醒或未通过项,安装前请自行复核。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