taiwan-md-knowledge-base
AI-native open knowledge base about Taiwan built with Astro v5, featuring bilingual content (zh-TW/en), D3.js knowledge graph, and structured Markdown SSOT architecture.
Best use case
taiwan-md-knowledge-base is best used when you need a repeatable AI agent workflow instead of a one-off prompt.
AI-native open knowledge base about Taiwan built with Astro v5, featuring bilingual content (zh-TW/en), D3.js knowledge graph, and structured Markdown SSOT architecture.
Teams using taiwan-md-knowledge-base should expect a more consistent output, faster repeated execution, less prompt rewriting.
When to use this skill
- You want a reusable workflow that can be run more than once with consistent structure.
When not to use this skill
- You only need a quick one-off answer and do not need a reusable workflow.
- You cannot install or maintain the underlying files, dependencies, or repository context.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/taiwan-md-knowledge-base/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How taiwan-md-knowledge-base Compares
| Feature / Agent | taiwan-md-knowledge-base | Standard Approach |
|---|---|---|
| Platform Support | Not specified | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
AI-native open knowledge base about Taiwan built with Astro v5, featuring bilingual content (zh-TW/en), D3.js knowledge graph, and structured Markdown SSOT architecture.
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
SKILL.md Source
# Taiwan.md Knowledge Base
> Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection.
Taiwan.md is an open-source, AI-native knowledge base about Taiwan built with Astro v5. It uses a Single Source of Truth (SSOT) architecture where all content lives in the `knowledge/` directory as Markdown files, and the website is a build-time projection. Features include bilingual support (Traditional Chinese as default + English), an interactive D3.js knowledge graph, and 96+ curated articles across 12 categories.
---
## Installation & Setup
### Prerequisites
- Node.js 18+
- npm or pnpm
### Clone and Install
```bash
git clone https://github.com/frank890417/taiwan-md.git
cd taiwan-md
npm install
```
### Development Server
```bash
npm run dev
# Site available at http://localhost:4321
```
### Build & Preview
```bash
npm run build
npm run preview
```
### Sync Knowledge to Content
```bash
bash scripts/sync.sh
# Copies knowledge/ → src/content/ for Astro build
```
---
## Project Architecture
```
taiwan-md/
├── knowledge/ ← SSOT: ALL content lives here
│ ├── History/ ← Chinese articles + _Hub.md
│ ├── Geography/
│ ├── Culture/
│ ├── Food/
│ ├── Art/
│ ├── Music/
│ ├── Technology/
│ ├── Nature/
│ ├── People/
│ ├── Society/
│ ├── Economy/
│ ├── Lifestyle/
│ ├── About/ ← Meta content
│ └── en/ ← English translations (mirrors zh-TW)
│ ├── History/
│ ├── Geography/
│ └── ...
├── scripts/
│ └── sync.sh ← Syncs knowledge/ → src/content/
├── src/
│ ├── pages/ ← Astro pages
│ ├── layouts/ ← Shared layouts
│ └── content/ ← Build-time projection (DO NOT EDIT)
├── public/
│ └── images/wiki/ ← Cached Wikimedia Commons images
└── docs/ ← Architecture & roadmap docs
```
**Critical rule:** Never edit files in `src/content/` directly. Always edit `knowledge/` and run `scripts/sync.sh`.
---
## Content Structure
### The 12 Categories
| Slug | Chinese | English |
|------|---------|---------|
| `history` | 歷史 | History |
| `geography` | 地理 | Geography |
| `culture` | 文化 | Culture |
| `food` | 美食 | Food |
| `art` | 藝術 | Art |
| `music` | 音樂 | Music |
| `technology` | 科技 | Technology |
| `nature` | 自然 | Nature |
| `people` | 人物 | People |
| `society` | 社會 | Society |
| `economy` | 經濟 | Economy |
| `lifestyle` | 生活 | Lifestyle |
### Article File Naming
```
knowledge/
├── Food/
│ ├── _Hub.md ← Category hub page (literary overview)
│ ├── bubble-tea.md ← Individual article (zh-TW)
│ └── beef-noodle.md
└── en/
└── Food/
├── _Hub.md ← English hub page
├── bubble-tea.md ← English translation
└── beef-noodle.md
```
---
## Writing Articles
### Chinese Article Template (`knowledge/[Category]/article-slug.md`)
```markdown
---
title: 珍珠奶茶
description: 台灣最具代表性的飲料文化,從夜市攤車到全球連鎖,珍珠奶茶如何征服世界。
category: food
date: 2024-01-15
tags: [飲食文化, 台灣之光, 夜市]
image: /images/wiki/bubble-tea-abc123.jpg
imageCaption: 台灣珍珠奶茶 | Wikimedia Commons | CC BY-SA 4.0
sources:
- title: 珍珠奶茶的起源考證
url: https://example.com/boba-origin
- title: 台灣飲料市場報告
url: https://example.com/beverage-report
---
## 30 秒認識
珍珠奶茶(波霸奶茶)誕生於 1980 年代台灣,現已成為全球年產值超過 30 億美元的飲料產業。
## 深度閱讀
### 起源爭議
台南翰林茶館與台中春水堂都聲稱是珍珠奶茶的發明者...
### 全球擴張
2010 年代,珍珠奶茶席捲歐美亞各大城市...
## 為什麼重要
珍珠奶茶不只是一杯飲料,它是台灣軟實力的最佳代言人——在沒有邦交的地方,台灣味道先到了。
## 參考資料
- [珍珠奶茶的起源考證](https://example.com/boba-origin)
- [台灣飲料市場報告](https://example.com/beverage-report)
```
### English Article Template (`knowledge/en/[Category]/article-slug.md`)
```markdown
---
title: Bubble Tea
description: Taiwan's most iconic beverage culture — how boba conquered the world from night market stalls to global chains.
category: food
date: 2024-01-15
tags: [food culture, taiwan pride, night market]
image: /images/wiki/bubble-tea-abc123.jpg
imageCaption: Taiwanese Bubble Tea | Wikimedia Commons | CC BY-SA 4.0
sources:
- title: Origins of Bubble Tea
url: https://example.com/boba-origin
- title: Taiwan Beverage Market Report
url: https://example.com/beverage-report
---
## 30-Second Overview
Bubble tea (boba) was born in 1980s Taiwan and has grown into a global industry worth over $3 billion annually.
## Deep Dive
### The Origin Debate
Both Hanlin Tea Room in Tainan and Chun Shui Tang in Taichung claim to have invented bubble tea...
### Global Expansion
In the 2010s, bubble tea swept across cities in Europe, America, and Asia...
## Why This Matters
Bubble tea isn't just a drink — it's Taiwan's finest soft power ambassador. Where there's no diplomatic recognition, Taiwanese flavor arrived first.
## References
- [Origins of Bubble Tea](https://example.com/boba-origin)
- [Taiwan Beverage Market Report](https://example.com/beverage-report)
```
### Hub Page Template (`knowledge/[Category]/_Hub.md`)
```markdown
---
title: 美食
titleEn: Food
description: 台灣的飲食文化是移民歷史、地理環境與創意精神的完美結晶。
category: food
---
## 關於這個分類
台灣是一個以食物說故事的地方...
## 精選文章
這個分類收錄了台灣飲食文化最具代表性的面向...
```
---
## Frontmatter Reference
### Required Fields
```yaml
---
title: "文章標題" # Display title
description: "一句話說明" # Meta description (150 chars max)
category: food # Must match one of 12 category slugs
date: 2024-01-15 # ISO date format
---
```
### Optional Fields
```yaml
---
tags: [tag1, tag2] # Array of tags for knowledge graph
image: /images/wiki/... # Must be from Wikimedia Commons cache
imageCaption: "..." # Attribution: Title | Source | License
sources: # REQUIRED: clickable URLs, no plain-text refs
- title: "Source Name"
url: https://...
---
```
---
## Adding Images (Wikimedia Commons Policy)
All images must be from Wikimedia Commons with verified CC licenses. Cache them locally:
```bash
# Download and cache a Wikimedia image
# Images are stored with MD5-hashed filenames
curl -o public/images/wiki/$(echo "filename.jpg" | md5sum | cut -d' ' -f1).jpg \
"https://commons.wikimedia.org/wiki/Special:FilePath/Taiwan_landscape.jpg"
```
Image attribution format in frontmatter:
```yaml
imageCaption: "Description | Wikimedia Commons | CC BY-SA 4.0"
```
---
## Knowledge Graph Integration
Articles automatically appear in the D3.js knowledge graph at `/graph`. Nodes are created from articles; edges are created from shared tags and cross-references.
### Linking Articles
Reference other articles within content using relative paths:
```markdown
台灣的[半導體產業](/technology/tsmc)是台積電...
See also: [Bubble Tea](/food/bubble-tea) for more on Taiwan's soft power.
```
### Tagging for Graph Connections
Use consistent tags to create knowledge graph bridges:
```markdown
# Both articles tagged [democratic transition] will be connected
tags: [democratic transition, civil society, 1990s]
```
---
## Sync Workflow
After editing any file in `knowledge/`, always sync before building:
```bash
# 1. Edit content
vim knowledge/Food/new-article.md
vim knowledge/en/Food/new-article.md
# 2. Sync to src/content/
bash scripts/sync.sh
# 3. Verify build
npm run build
# 4. Preview
npm run preview
```
---
## Three-Layer Depth Pattern
Every article should follow this structure for AI-readability and varied reading levels:
```markdown
## 30 秒認識 / 30-Second Overview
[2-3 sentences, the essential fact]
## 深度閱讀 / Deep Dive
### Subsection 1
[Detailed exploration with data]
### Subsection 2
[Historical context or comparison]
## 為什麼重要 / Why This Matters
[Curatorial perspective — answer "why should the world care?"]
## 參考資料 / References
[Clickable URLs only — no plain-text citations]
```
---
## Contributing via PR
### Full PR Workflow
```bash
# 1. Fork and clone
git clone https://github.com/YOUR_USERNAME/taiwan-md.git
cd taiwan-md
# 2. Create branch
git checkout -b add/food/scallion-pancake
# 3. Add zh-TW article
cat > knowledge/Food/scallion-pancake.md << 'EOF'
---
title: 蔥抓餅
description: ...
category: food
date: 2024-01-20
sources:
- title: Source
url: https://...
---
Content here...
EOF
# 4. Add English translation
mkdir -p knowledge/en/Food
cat > knowledge/en/Food/scallion-pancake.md << 'EOF'
---
title: Scallion Pancake
...
EOF
# 5. Sync and test
bash scripts/sync.sh
npm run build
# 6. Commit and PR
git add knowledge/
git commit -m "feat(food): add scallion pancake article (zh+en)"
git push origin add/food/scallion-pancake
```
### Commit Message Convention
```
feat(category): add [article name] article (zh+en)
fix(category): correct [article name] factual error
i18n(category): add English translation for [article name]
feat(graph): add knowledge graph connections for [topic]
```
---
## AI-Native Features
### llms.txt
The site exposes `/llms.txt` for AI consumption. When writing content, use structured headers that AI can parse:
```markdown
# Title
**Key fact:** One-sentence essential truth.
## Context
...
## Significance
...
```
### Meta AI Summary Tag
Pages include `<meta ai-summary>` — write descriptions that work as standalone AI context:
```yaml
description: "台積電(TSMC)生產全球90%最先進晶片,是台灣的「矽盾」——台灣的地緣政治生存策略。"
```
---
## Astro Page Patterns
### Category Page (`src/pages/[category].astro`)
```astro
---
import { getCollection } from 'astro:content';
const category = 'food';
const articles = await getCollection('knowledge', ({ data }) =>
data.category === category
);
---
<ul>
{articles.map(article => (
<li>
<a href={`/${category}/${article.slug}`}>{article.data.title}</a>
<p>{article.data.description}</p>
</li>
))}
</ul>
```
### Bilingual Route Pattern
```
/food/bubble-tea ← zh-TW (default)
/en/food/bubble-tea ← English
```
---
## Content Quality Checklist
Before submitting a PR, verify:
- [ ] Both `knowledge/[Category]/article.md` (zh-TW) and `knowledge/en/[Category]/article.md` (en) exist
- [ ] All `sources` entries have clickable `url` fields (no plain-text-only references)
- [ ] Article follows three-layer depth: 30-sec → deep dive → why it matters
- [ ] Images sourced from Wikimedia Commons with proper `imageCaption` attribution
- [ ] `category` slug matches one of the 12 valid categories exactly
- [ ] `bash scripts/sync.sh && npm run build` completes without errors
- [ ] Factual claims are verified against cited sources
---
## Troubleshooting
### Build fails after adding article
```bash
# Check frontmatter syntax
cat knowledge/Food/my-article.md | head -20
# Common issue: missing required fields
# Ensure title, description, category, date are all present
# Re-sync and rebuild
bash scripts/sync.sh
npm run build 2>&1 | grep ERROR
```
### Article not appearing in knowledge graph
```bash
# Ensure tags array is populated
# Check category slug matches exactly (case-sensitive)
# Verify sync was run after editing
bash scripts/sync.sh
```
### English article not showing at /en/...
```bash
# Verify file exists at correct path
ls knowledge/en/Food/my-article.md
# Check category field matches zh-TW article exactly
grep "category:" knowledge/Food/my-article.md
grep "category:" knowledge/en/Food/my-article.md
```
### Images not loading
```bash
# Images must be cached locally in public/images/wiki/
ls public/images/wiki/
# Verify frontmatter path starts with /images/wiki/
grep "image:" knowledge/Food/my-article.md
# Should be: image: /images/wiki/filename-hash.jpg
```
### Sync script permission error
```bash
chmod +x scripts/sync.sh
bash scripts/sync.sh
```
---
## Resources
- **Live site:** https://taiwan.md
- **Knowledge graph:** https://taiwan.md/graph
- **Contribute guide:** https://taiwan.md/contribute
- **llms.txt:** https://taiwan.md/llms.txt
- **License:** CC BY-SA 4.0 (content) + MIT (code)
- **Contact:** cheyu.wu@monoame.comRelated Skills
understand-anything-knowledge-graph
Turn any codebase into an interactive knowledge graph using Claude Code skills — explore, search, and ask questions about any project visually.
taiwan-equity-research-coverage
Structured equity research database for 1,735 Taiwan-listed companies with wikilink knowledge graph, supply chain mapping, and financial data tools.
openviking-context-database
Expert skill for using OpenViking, the open-source context database for AI Agents that manages memory, resources, and skills via a filesystem paradigm.
dbx-database-client
Open-source lightweight cross-platform database management tool built with Tauri, Vue 3, and Rust supporting MySQL, PostgreSQL, SQLite, Redis, MongoDB, DuckDB, ClickHouse, and SQL Server.
cabinet-ai-knowledge-base
AI-first knowledge base and startup OS with file-based storage, AI agents, scheduled jobs, and embedded apps
```markdown
---
zeroboot-vm-sandbox
Sub-millisecond VM sandboxes for AI agents using copy-on-write KVM forking via Zeroboot
yourvpndead-vpn-detection
Android app that detects VPN/proxy servers (VLESS/xray/sing-box) via local SOCKS5 vulnerability, exposing exit IPs and server configs without root
xata-postgres-platform
Expert skill for Xata open-source cloud-native Postgres platform with copy-on-write branching, scale-to-zero, and Kubernetes deployment
x-mentor-skill-nuwa
AI-powered X (Twitter) content strategy skill that distills methodologies from 6 top creators + open-source algorithm data into actionable writing, growth, and monetization guidance.
wx-favorites-report
End-to-end pipeline to extract, decrypt, and visualize WeChat Mac favorites from encrypted SQLite DB into an interactive HTML report.
wterm-web-terminal
Web terminal emulator with Zig/WASM core, DOM rendering, and React/vanilla JS bindings