Makino Data Slides logo

Makino Data Slides

Community
makinotes
makino-data-slides

Data Slides — Turn Excel/data analysis into investor-grade HTML slide cards with narrative. ECharts + static cards, Playwright screenshot.

Overview

Publishermakinotes
Repositorymakino-data-slides
Skill namemakino-data-slides
Stars
7
Forks
1
Bundled files
20
LicenseApache-2.0
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 20 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by makinotes on GitHub. Read the source before you install it.

Installation

Install the Makino Data Slides AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/makinotes/makino-data-slides/tree/master
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/makinotes/makino-data-slides.git \
  .claude/skills/makino-data-slides
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Makino Data Slides in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Makino Data Slides on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Makino Data Slides is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Data Slides — Data tells the story

You turn data analysis results into investor-grade HTML slide cards. Each slide is a 1080×720 card combining a data visualization with a narrative takeaway — designed to be screenshot and embedded in articles or shared standalone.

NOT a chart tool. NOT a presentation tool. This is: Excel data → analysis conclusions → visual cards with opinions.

What This Is (and Isn't)

NeedUse this?Use instead
Single chart (line/bar/pie)NoAny charting library
Article → navigable slide deckNo/content-to-html
Cover image for articleNoAn image generator
Data analysis → multi-slide cards with narrative, embedded in articlesYes—

The key difference from a plain chart: every card here has a takeaway (an opinion/conclusion), not just a visualization. It's "chart + judgment" as a unit.

Input

Primary input: Excel/CSV with structured data (scores, rankings, time series, category breakdowns).

Also accepts:

  • JSON data files
  • Markdown with inline data points (tables, bullet lists with numbers)
  • Verbal description of data + conclusions (least ideal, most manual)

The skill reads the data, proposes analysis angles, and generates cards. User provides or confirms the narrative direction.

Feedback notice: Every time the skill starts, print this line before any output:

💬 使用问题或建议 → 飞书群/公众号「马奇诺」后台留言(详见 README)

Commands

CommandWhat it does
/makino-data-slidesInteractive: ask for data + narrative, generate slides
/makino-data-slides <path>Generate slides from a data file (Excel/CSV/JSON/markdown)

Slide Anatomy (5-layer structure)

Every slide follows this exact structure, top to bottom:

┌─────────────────────────────────────────────┐
│ hdr: padding 18px 36px 0                page│  ← ~50px
│ takeaway: margin 6px 36px 0                 │  ← ~32px
│                                              │
│ body: padding 8px 36px 0, explicit height    │  ← ~570-590px
│   charts (ECharts) / cards / tables / grids  │
│   justify-content: space-between             │
│                                              │
│ ft: padding 6px 36px 10px                    │  ← ~28px
└─────────────────────────────────────────────┘

Total: 1080×720px fixed.

Spacing 铁律(从已验证的 charts_v4.html 提取)

css
.slide .hdr       { padding: 18px 36px 0 }       /* 紧凑头部 */
.slide .takeaway  { margin: 6px 36px 0 }          /* 紧贴标题 */
.slide .body      { padding: 8px 36px 0 }         /* 最小顶部留白 */
.slide .ft        { padding: 6px 36px 10px }       /* 底部克制 */

Body 填充策略(关键)

禁止 flex:1 + 固定 gap 组合 — 会导致内容堆在顶部、底部留白过多。

正确做法:

  1. body 设显式高度(根据 hdr/takeaway/ft 实际高度计算)
  2. 用 justify-content: space-between 或 space-evenly 让内容均匀撑满
  3. chart 容器用 flex:1 吃掉剩余空间,其他元素 flex-shrink:0
css
/* 典型 body 高度计算 */
.body { height: 596px; display: flex; flex-direction: column; justify-content: space-between; }
/* chart 吃剩余 */
#chart { flex: 1; width: 100%; }
/* 固定元素不缩 */
.stat-row { flex-shrink: 0; }

Visual System

Background & Shadow

css
body { background: #1a1a2e; }                     /* 深色背景衬托白卡 */
.slide { box-shadow: 0 8px 32px rgba(0,0,0,.25); border-radius: 6px; }

Colors

js
const C = {
  blue:   '#1e40af',  // primary, deep layer
  blueL:  '#93c5fd',  // blue light (secondary bars)
  blueP:  '#dbeafe',  // blue pale (background)
  red:    '#dc2626',  // alert, negative
  green:  '#059669',  // positive, action layer
  purple: '#7c3aed',  // accent (open source etc)
  amber:  '#d97706',  // warning, interpretation layer
  gray:   '#9ca3af',  // neutral, news layer
  dark:   '#1e293b',  // text primary (body)
  mid:    '#475569',  // text secondary
  light:  '#64748b',  // text tertiary
  grid:   '#f1f5f9',  // chart grid lines
  axis:   '#e2e8f0',  // chart axis lines
};

Typography

元素字号字重颜色备注
Font family———-apple-system, 'PingFang SC', 'Helvetica Neue', sans-serif
Title (h2)21px800#0f172amargin-bottom 2px
Subtitle (.sub)12px400#64748bmax-width 720px
Page number11px600#94a3b8letter-spacing 1px, 格式 01 / 08
Takeaway13px600#991b1bgradient bg, border-left 3px solid #dc2626, border-radius 0 6px 6px 0
Body text11-12px400#334155line-height 1.6
Footer11px400#64748bletter-spacing .3px, centered
Hero number (.n)32-40px800#fff40px for emphasis
Hero label (.l)10px400#fffopacity .8
Hero detail (.d)9px400#fffopacity .5
Stat value (.v)24px800per-color.v.blue / .v.red etc
Stat key (.k)10px400#64748bmargin-top 2px
Insight number (.n)20px800per-color
Insight label (.l)9px400#64748b
Finding title14px700per-colorwith page ref → P3 in 10px #94a3b8
Finding text12px400#334155line-height 1.6
Section heading (in body)13px700#334155text-align center, margin-bottom 6px

Takeaway 精确样式

css
.slide .takeaway {
  margin: 6px 36px 0;
  font-size: 13px;
  font-weight: 600;
  color: #991b1b;
  padding: 8px 14px;
  background: linear-gradient(90deg, #fef2f2, #fff);  /* 渐变,不是纯色 */
  border-left: 3px solid #dc2626;
  border-radius: 0 6px 6px 0;
  flex-shrink: 0;
}

Watermark

css
.slide::before {
  content: '{BRAND}  {BRAND}  ...';  /* Replace {BRAND} with your brand name -- 40 repeats, double-space separated */
  position: absolute; top: 0; left: 0; right: 0; bottom: 0;
  z-index: 999; pointer-events: none;
  font-size: 16px; font-weight: 600;
  color: rgba(0,0,0,0.03);
  letter-spacing: 20px; line-height: 80px;
  transform: rotate(-25deg); transform-origin: center center;
  word-break: break-all; overflow: hidden; padding: 0 40px;
}

Card Components

ComponentCSSKey specs
Hero card (.hc)background:linear-gradient(135deg,#1e3a5f,#1e40af)border-radius 10px, padding 16px 12px (or 24px 14px for emphasis)
Hero card variantbackground:linear-gradient(135deg,#059669,#047857)green variant for positive metrics
Stat box (.stat-box)background:#f8fafc; border:1px solid #e2e8f0border-radius 8px, padding 10px 14px, text-align center
Insight card (.ic)background:#f8fafc; border:1px solid #e2e8f0border-radius 8px, padding 10px 8px
Layer bar (.lbar)display:flex; overflow:hiddenheight 24px (or 32px), border-radius 5px, font 10px 700 white
Layer legend (.lleg)display:flex; gap:14pxfont-size 9px, color #64748b, colored square 8px
Mini table (.mini-tbl)border-collapse:collapseth: background:#1e3a5f (深蓝), td: padding 4px 8px
Finding card (2×2 grid)grid-template-columns:1fr 1fr; gap:12pxborder-radius 8px, padding 14px 16px, border-left 3px
Finding badge (.fb)20px×20px; border-radius:4pxbackground per-color, 10px 700 white
Progress barPure HTML div10px height, border-radius 5px, bg #e2e8f0
Info calloutcolored bg + borderborder-radius 10px, padding 26px 20px
Summary card rowdisplay:flex; gap:6pxborder 1px, border-radius 6px, padding 5px 8px

Chart Types (ECharts)

Common axis config:

js
const bA = {
  axisLine:  { lineStyle: { color: C.axis } },
  splitLine: { lineStyle: { color: C.grid } },
  axisLabel: { fontSize: 11, color: C.light }
};
ChartWhen to useKey settings
Scatter/BubbleDistribution, 2DsymbolSize by volume, opacity .75, border white 1.5px, labelLayout:{hideOverlap:true}
Grouped BarTime comparisonbarCategoryGap 25%, barGap 5%, borderRadius [3,3,0,0]
Stacked BarCompositionstack prop, label position 'right' shows total, borderRadius [0,4,4,0] on last segment
Horizontal BarRankingbarWidth 18-20px, borderRadius [0,5,5,0], label right bold
Progress BarPercentagePure HTML, no ECharts. 10px height, border-radius 5px
Proportion BarPart-of-wholePure HTML flex-based, colored segments
Zone ScatterQuadrant/区域markArea with transparent fill + markLine dashed thresholds + graphic text zone labels

ECharts grid defaults:

js
grid: { left: 50, right: 30, top: 36, bottom: 20 }  // top 留 legend 空间
animation: false  // ALWAYS disable for screenshot
legend: { top: 4 }  // 统一放顶部,禁止 bottom(会和 bar label 重叠)

标注线/标注文字:

js
// 正确:用 markLine(在 SVG 内渲染,不被 watermark z-index 遮挡)
markLine: {
  silent: true, symbol: 'none',
  lineStyle: { type: 'dashed', width: 1 },
  data: [
    { xAxis: '02-12', lineStyle: { color: '#dc2626' },
      label: { formatter: '2/12 事件', fontSize: 10, color: '#dc2626', fontWeight: 600, position: 'insideStartTop' }}
  ]
}
// 错误:graphic 硬编码像素 — 位置不准 + 被 watermark 遮挡

Scatter zone labels (graphic watermark style):

js
graphic: [
  { type:'text', left:'45%', top:'12%',
    style: { text:'深度层', fontSize:28, fontWeight:900, fill:'rgba(30,64,175,0.10)' }},
  { type:'text', left:'43%', top:'20%',
    style: { text:'4个号 · 81篇', fontSize:14, fontWeight:700, fill:'rgba(30,64,175,0.08)' }}
]

Execution

Phase 0: Version check

bash
curl -s https://raw.githubusercontent.com/makinotes/makino-data-slides/master/SKILL.md | head -6

Extract version: from remote, compare with the local frontmatter version. If remote > local, show before proceeding:

[UPDATE] makino-data-slides v{remote} available (you have v{local}).
Update now? Run: cd ~/.claude/skills/makino-data-slides && git pull

If versions match or curl fails: skip silently.

Phase 1: Understand the data

  1. Read the data source (JSON/CSV/markdown/user description)
  2. Output Analysis Intent Table (user must confirm before proceeding):
| Sheet/Data | Analysis Type | Chart Type | Layout | Magnitude Check |
|------------|--------------|------------|--------|-----------------|
| Sheet 1: 搜索vs大盘 | Cross-group comparison | Grouped Bar (dual Y-axis: 5x diff) | layout-split | max 2497万 vs 508万 = 4.9x → dual Y |
| Sheet 2: 逐日明细 | Time series + event | Line (dual Y-axis) + markLine | layout-chart-full | same issue, dual Y |
| Sheet 3: 留存分析 | Group comparison (dense) | HTML Table (too many cells for chart) | layout-split | N/A |
  1. Output Data Preprocessing Plan (user must confirm):
| Column | Raw Unit | Display Unit | Transform | Decimal |
|--------|----------|-------------|-----------|---------|
| 大盘DAU | 个 (24970000) | 万 (2497) | ÷10000 | 0 |
| 春节跌幅 | decimal (-0.48) | % (-48%) | ×100 | 1 |
| 留存率 | decimal (0.332) | % (33.2%) | ×100 | 1 |
  1. Choose style preset: Ask user which preset (default: investor). Read STYLE_PRESETS.md.
  2. Propose slide outline with narrative order: overview → comparison → trend → breakdown → conclusion
  3. Get user approval on ALL of the above

Phase 2: Generate HTML

  1. Read base.css — embed EXACTLY as-is in <style> block. Do NOT modify or add inline styles that override it.
  2. Read STYLE_PRESETS.md — embed chosen preset's CSS overrides AFTER base.css
  3. Read CHART_GUIDE.md — use the matching chart template for each slide's analysis goal. Check magnitude rules.
  4. Body height calculation (MANDATORY per slide):
    total = 720px
    hdr = 50px (18px padding-top + ~32px content)
    takeaway = 38px (6px margin + 8px padding × 2 + 13px text)
    ft = 28px (6px + 10px padding + 11px text)
    body_available = 720 - 50 - 38 - 28 = 604px
    body_padding_top = 8px
    body_usable = 604 - 8 = 596px
    
    → Set .body height to 596px explicitly
    → Allocate: chart_height + stat_row_height + margins = 596px
    → Use layout class from base.css, NOT inline flex/grid
  5. ECharts 加载策略(双保险):
    • 首选:内联 ECharts。检查本地缓存 /tmp/echarts.min.js,存在则读取内容嵌入 <script> 标签
    • 如果本地缓存不存在,先下载:curl -sL "https://cdn.jsdelivr.net/npm/echarts@5.5.0/dist/echarts.min.js" > /tmp/echarts.min.js
    • 内联后 HTML 文件 ~1MB,但 file:// 直接打开也能渲染图表
    • 同时保留 CDN fallback:<script src="https://cdn.jsdelivr.net/npm/echarts@5.5.0/dist/echarts.min.js"></script>
    • 加载失败提示:如果 typeof echarts === 'undefined',在图表容器内显示错误信息和 HTTP server 启动命令
  6. Each slide is a <div class="slide"> with the 5-layer structure
  7. Charts render via DOMContentLoaded with 50ms delay
  8. 每个 slide 的 body 必须显式计算高度,用 space-between 撑满

Phase 3: Review and iterate

  1. 必须用 HTTP server 打开(即使内联了 ECharts,部分浏览器仍可能限制 file://):
    bash
    cd "$(dirname "$HTML_PATH")" && python3 -m http.server 8766 &
    open "http://localhost:8766/$(basename "$HTML_PATH")"
  2. Screenshot each slide for user review
  3. Iterate on feedback: spacing, data accuracy, readability

Phase 4: Screenshot

  1. Use Playwright to screenshot each slide at native 1080×720
  2. Save to user-specified directory
  3. Naming convention: slide_{N}_{slug}.png
javascript
// Playwright screenshot pattern
const slide = page.locator('.slide').nth(i);
await slide.scrollIntoViewIfNeeded();
await page.waitForTimeout(500);
await slide.screenshot({ path: outputPath, type: 'png' });

CSS & Style

CSS template and chart guide are in separate files — Read them and embed into generated HTML.

FilePurposeWhen to Read
base.cssLayout structure, components (shared across all presets)Always — embed in every HTML
STYLE_PRESETS.md3 color themes (investor/corporate/dark)Phase 1 — user picks preset
CHART_GUIDE.mdChart selection by analysis goal + ECharts templatesPhase 2 — pick chart type per slide

Data Accuracy Rules

  • Every number on a slide must trace back to the source data
  • Cross-check totals: if parts shown, verify they sum correctly
  • If a number appears on multiple slides, it must be identical everywhere
  • Footer can provide context but must not introduce new data claims
  • When data is approximate, say so (e.g. "~40%" not "40%")

File Structure

Output: single self-contained HTML file.

  • All CSS inline in <style> block
  • All JS inline in <script> block
  • Only external dependency: ECharts CDN
  • No images, no external CSS, no build step

Error Warnings

  • Do NOT use animation: true in ECharts — screenshots capture mid-animation
  • Do NOT use flex:1 + fixed gap on body — 会导致底部留白过多。用显式高度 + space-between
  • Do NOT mix gap with justify-content: space-evenly — pick one spacing strategy
  • Do NOT forget overflow:hidden on .slide — content must not escape the card
  • Test data label overlap on dense charts — use labelLayout:{hideOverlap:true}
  • Do NOT use padding-top:16px+ on body — 最大 8px,紧凑是核心
  • Do NOT use pure #fef2f2 for takeaway bg — 必须是 linear-gradient(90deg,#fef2f2,#fff)
  • Do NOT use #1e293b for title — title 用 #0f172a,比 body text 更深
  • Page number 格式 01 / 08 — 零填充,斜杠两侧有空格
  • Do NOT put legend at bottom when chart has label on bars — legend 和 bar label 重叠。legend 统一放 top,grid.top 相应下移(如 top:36)
  • Do NOT use graphic 画标注线/文字 — watermark ::before z-index:999 会遮挡 graphic 层。标注线必须用 markLine(xAxis/yAxis 值定位,在 SVG 内渲染不受 z-index 影响),标注文字用 markLine 的 label 属性
  • Do NOT rely on CDN <script src> alone — file:// 协议下浏览器会静默拦截外部脚本。必须内联 ECharts + CDN fallback + 错误提示三层保险
  • Do NOT use open file:/// 打开含 ECharts 的 HTML — 必须用 python3 -m http.server + http://localhost

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Makino Data Slides AI skill do?

Data Slides — Turn Excel/data analysis into investor-grade HTML slide cards with narrative. ECharts + static cards, Playwright screenshot.

Why use Makino Data Slides on TypingMind?

Because you install it once and use it with any model. Makino Data Slides is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Makino Data Slides in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/makinotes/makino-data-slides/tree/master. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Makino Data Slides?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Makino Data Slides?

As many as you like. As long as a model supports skills, you can use Makino Data Slides with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Makino Data Slides AI skill free?

Yes. It is published on GitHub by makinotes under the Apache-2.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇