tapestry-skills-for-claude-code

内容来源：README.md（说明文档） · 原始地址 · 查看安装指南

原始内容

# Tapestry Skills for Claude Code

A collection of productivity skills for [Claude Code](https://claude.com/claude-code) that help you work faster and learn better.

**Tapestry weaves learning content into action.** Extract any content (YouTube, articles, PDFs) and automatically create implementation plans.

## Requirements

- **[UV](https://docs.astral.sh/uv/)** - Fast Python package manager (required)
- **Python 3.10+** - UV will manage this for you

Install UV:
```bash
curl -LsSf https://astral.sh/uv/install.sh | sh
# or
brew install uv
```

## Skills Included

### 0. Tapestry (Master Skill)
The unified workflow that orchestrates everything. Just say `tapestry <URL>` and it:
1. Detects content type (YouTube, article, PDF)
2. Extracts clean content
3. Automatically creates a Ship-Learn-Next action plan

**One command. Complete workflow. From learning to shipping.**

### 1. YouTube Transcript Downloader
Download and clean YouTube video transcripts with automatic deduplication and readable formatting.

**Features:**
- Downloads transcripts using yt-dlp (pinned version via UV)
- Removes duplicate lines (YouTube VTT format issue)
- Uses video title for filename
- Fallback to Whisper transcription if no subtitles available
- Automatic cleanup of temporary files

### 2. Article Extractor
Extract clean, readable content from web articles and blog posts, removing ads, navigation, and clutter.

**Features:**
- Uses Mozilla Readability or trafilatura for extraction
- Removes ads, navigation, and newsletter signups
- Saves as clean plain text
- Uses article title for filename
- Built-in fallback extractor when external tools unavailable

### 3. Ship-Learn-Next Action Planner
Transform passive learning content (transcripts, articles, tutorials) into actionable implementation plans using the Ship-Learn-Next framework.

**Features:**
- Converts advice into concrete, shippable iterations (reps)
- Creates 5-rep action plans with timelines
- Focuses on DOING over studying
- Includes reflection questions after each rep

## Installation

### Plugin Install (Recommended)

```bash
# Install directly as a Claude Code plugin
claude /install-plugin https://github.com/ryanhudson/tapestry-skills-for-claude-code
```

### Script Install (Alternative)

```bash
# Clone the repository
git clone https://github.com/ryanhudson/tapestry-skills-for-claude-code.git

# Run the installation script
cd tapestry-skills-for-claude-code
chmod +x install.sh
./install.sh
```

The installer will:
1. Check for UV (offer to install if missing)
2. Verify Python 3.10+
3. Install all dependencies via `uv sync`
4. Verify the utilities work
5. Symlink skills to `~/.claude/skills/`

### Manual Installation

```bash
# 1. Clone and enter directory
git clone https://github.com/yourusername/tapestry-skills-for-claude-code.git
cd tapestry-skills-for-claude-code

# 2. Install dependencies
uv sync

# 3. Verify installation
uv run tapestry-validate-url https://example.com
uv run tapestry-sanitize-filename "Test: File/Name"

# 4. Symlink skills to Claude (or use plugin install)
mkdir -p ~/.claude/skills
ln -sfn "$(pwd)/skills/tapestry" ~/.claude/skills/tapestry
ln -sfn "$(pwd)/skills/youtube-transcript" ~/.claude/skills/youtube-transcript
ln -sfn "$(pwd)/skills/article-extractor" ~/.claude/skills/article-extractor
ln -sfn "$(pwd)/skills/ship-learn-next" ~/.claude/skills/ship-learn-next

# OR install as a plugin:
# claude /install-plugin https://github.com/ryanhudson/tapestry-skills-for-claude-code
```

## Usage

### Tapestry (Recommended - Use This!)

The simplest way to use Tapestry skills. One command extracts content and creates your action plan:

```
"tapestry https://www.youtube.com/watch?v=VIDEO_ID"
"weave https://example.com/article"
"help me plan https://example.com/paper.pdf"
"make this actionable https://blog.com/post"
```

**All these phrases work**: tapestry, weave, help me plan, extract and plan, make this actionable

The skill will:
1. Detect content type (YouTube/article/PDF)
2. Extract clean content
3. Create a Ship-Learn-Next action plan automatically
4. Save both files
5. Ask: "When will you ship Rep 1?"

### YouTube Transcript Downloader

```
"Download the transcript for https://www.youtube.com/watch?v=VIDEO_ID"
```

### Article Extractor

```
"Extract the article from https://example.com/blog-post"
"Download this article without the ads"
```

### Ship-Learn-Next Action Planner

```
"Turn this transcript into an implementation plan"
"Make this actionable using the Ship-Learn-Next framework"
```

## Built-in Utilities

All utilities are available via `uv run` from the project root:

| Utility | Description |
|---------|-------------|
| `tapestry-validate-url` | URL validation with SSRF protection |
| `tapestry-sanitize-filename` | Safe filename generation |
| `tapestry-safe-download` | Secure file downloads with size limits |
| `tapestry-vtt-to-text` | VTT subtitle to plain text conversion |
| `tapestry-extract-html` | Fallback HTML content extractor |

Example:
```bash
uv run tapestry-validate-url "https://example.com"
uv run tapestry-sanitize-filename "My: Unsafe/Filename"
uv run tapestry-vtt-to-text captions.vtt --output transcript.txt
```

## Dependencies

All Python dependencies are managed via UV and pinned in `pyproject.toml`:

| Package | Purpose |
|---------|---------|
| `yt-dlp` | YouTube video/transcript downloads |
| `trafilatura` | Article extraction |
| `openai-whisper` (optional) | Transcription for videos without subtitles |

**Optional system tools** (install separately for best results):
- `reader` - Mozilla Readability CLI (`npm install -g reader-cli`)
- `pdftotext` - PDF text extraction (`brew install poppler`)

## Examples

### Example 1: Tapestry Unified Workflow

```
User: "tapestry https://www.youtube.com/watch?v=dQw4w9WgXcQ"

Claude:
Tapestry Workflow Starting...
Detected: youtube
Extracting YouTube transcript...
Saved transcript: Never Gonna Give You Up.txt

Creating action plan...
Quest: Master Video Production Techniques
Saved plan: Ship-Learn-Next Plan - Master Video Production.md

Tapestry Complete!
Content: Never Gonna Give You Up.txt
Plan: Ship-Learn-Next Plan - Master Video Production.md

Rep 1 (This Week): Film and edit a 60-second video
When will you ship Rep 1?
```

### Example 2: Download YouTube Transcript

```
User: "Download transcript for https://www.youtube.com/watch?v=VIDEO_ID"

Claude:
Checked available subtitles
Downloaded auto-generated transcript
Converted to readable format
Removed duplicate lines
Saved to: Video Title.txt
```

### Example 3: Extract Article

```
User: "Extract https://example.com/how-to-build-saas"

Claude:
Using trafilatura for extraction
Extracted article: How to Build a SaaS in 30 Days
Saved to: How to Build a SaaS in 30 Days.txt

Preview (first 10 lines):
[Clean article text without ads or navigation...]
```

## Philosophy

These skills are built on the principle that **learning = doing better, not knowing more**.

### Ship-Learn-Next Framework
- **Ship**: Create something real (code, content, product)
- **Learn**: Honest reflection on what happened
- **Next**: Plan the next iteration based on learnings

100 reps beats 100 hours of study.

## Project Structure

```
tapestry-skills-for-claude-code/
├── plugin.json             # Claude Code plugin manifest
├── pyproject.toml          # Dependencies (UV/pip)
├── src/tapestry/           # Python utilities
│   ├── validate_url.py
│   ├── sanitize_filename.py
│   ├── safe_download.py
│   ├── vtt_to_text.py
│   └── html_extractor.py
├── skills/                 # Claude Code skills
│   ├── tapestry/           # Master skill
│   ├── youtube-transcript/ # YouTube extraction skill
│   ├── article-extractor/  # Article extraction skill
│   └── ship-learn-next/    # Action planning skill
├── tests/                  # Test suite
├── shared/references/      # Security documentation
└── install.sh              # Installation script
```

## Contributing

Found a bug or want to add a feature? Contributions are welcome!

1. Fork the repository
2. Create your feature branch (`git checkout -b feature/amazing-skill`)
3. Commit your changes (`git commit -m 'Add amazing skill'`)
4. Push to the branch (`git push origin feature/amazing-skill`)
5. Open a Pull Request

## License

MIT License - see [LICENSE](https://github.com/ryanhudson/tapestry-skills-for-claude-code/tree/HEAD/LICENSE) file for details

## Acknowledgments

- **UV**: Fast Python package management from Astral
- **Ship-Learn-Next framework**: Inspired by the ShipLearnNext GPT methodology
- **yt-dlp**: Excellent tool for downloading YouTube content
- **OpenAI Whisper**: State-of-the-art speech recognition
- **Mozilla Readability**: Clean article extraction algorithm
- **trafilatura**: Python web scraping and content extraction

## Support

Having issues? Please [open an issue](https://github.com/yourusername/tapestry-skills-for-claude-code/issues) on GitHub.

---

**Made with Claude Code**

Learn more about Claude Code at [claude.com/claude-code](https://claude.com/claude-code)

Tapestry Skills for Claude Code

A collection of productivity skills for Claude Code that help you work faster and learn better.

Tapestry weaves learning content into action. Extract any content (YouTube, articles, PDFs) and automatically create implementation plans.

Requirements

UV - Fast Python package manager (required)
Python 3.10+ - UV will manage this for you

Install UV:

curl -LsSf https://astral.sh/uv/install.sh | sh
# or
brew install uv

Skills Included

0. Tapestry (Master Skill)

The unified workflow that orchestrates everything. Just say tapestry <URL> and it:

Detects content type (YouTube, article, PDF)
Extracts clean content
Automatically creates a Ship-Learn-Next action plan

One command. Complete workflow. From learning to shipping.

1. YouTube Transcript Downloader

Download and clean YouTube video transcripts with automatic deduplication and readable formatting.

Features:

Downloads transcripts using yt-dlp (pinned version via UV)
Removes duplicate lines (YouTube VTT format issue)
Uses video title for filename
Fallback to Whisper transcription if no subtitles available
Automatic cleanup of temporary files

2. Article Extractor

Extract clean, readable content from web articles and blog posts, removing ads, navigation, and clutter.

Features:

Uses Mozilla Readability or trafilatura for extraction
Removes ads, navigation, and newsletter signups
Saves as clean plain text
Uses article title for filename
Built-in fallback extractor when external tools unavailable

3. Ship-Learn-Next Action Planner

Transform passive learning content (transcripts, articles, tutorials) into actionable implementation plans using the Ship-Learn-Next framework.

Features:

Converts advice into concrete, shippable iterations (reps)
Creates 5-rep action plans with timelines
Focuses on DOING over studying
Includes reflection questions after each rep

Installation

Plugin Install (Recommended)

# Install directly as a Claude Code plugin
claude /install-plugin https://github.com/ryanhudson/tapestry-skills-for-claude-code

Script Install (Alternative)

# Clone the repository
git clone https://github.com/ryanhudson/tapestry-skills-for-claude-code.git

# Run the installation script
cd tapestry-skills-for-claude-code
chmod +x install.sh
./install.sh

The installer will:

Check for UV (offer to install if missing)
Verify Python 3.10+
Install all dependencies via uv sync
Verify the utilities work
Symlink skills to ~/.claude/skills/

Manual Installation

# 1. Clone and enter directory
git clone https://github.com/yourusername/tapestry-skills-for-claude-code.git
cd tapestry-skills-for-claude-code

# 2. Install dependencies
uv sync

# 3. Verify installation
uv run tapestry-validate-url https://example.com
uv run tapestry-sanitize-filename "Test: File/Name"

# 4. Symlink skills to Claude (or use plugin install)
mkdir -p ~/.claude/skills
ln -sfn "$(pwd)/skills/tapestry" ~/.claude/skills/tapestry
ln -sfn "$(pwd)/skills/youtube-transcript" ~/.claude/skills/youtube-transcript
ln -sfn "$(pwd)/skills/article-extractor" ~/.claude/skills/article-extractor
ln -sfn "$(pwd)/skills/ship-learn-next" ~/.claude/skills/ship-learn-next

# OR install as a plugin:
# claude /install-plugin https://github.com/ryanhudson/tapestry-skills-for-claude-code

Usage

Tapestry (Recommended - Use This!)

The simplest way to use Tapestry skills. One command extracts content and creates your action plan:

"tapestry https://www.youtube.com/watch?v=VIDEO_ID"
"weave https://example.com/article"
"help me plan https://example.com/paper.pdf"
"make this actionable https://blog.com/post"

All these phrases work: tapestry, weave, help me plan, extract and plan, make this actionable

The skill will:

Detect content type (YouTube/article/PDF)
Extract clean content
Create a Ship-Learn-Next action plan automatically
Save both files
Ask: "When will you ship Rep 1?"

YouTube Transcript Downloader

"Download the transcript for https://www.youtube.com/watch?v=VIDEO_ID"

Article Extractor

"Extract the article from https://example.com/blog-post"
"Download this article without the ads"

Ship-Learn-Next Action Planner

"Turn this transcript into an implementation plan"
"Make this actionable using the Ship-Learn-Next framework"

Built-in Utilities

All utilities are available via uv run from the project root:

Utility	Description
`tapestry-validate-url`	URL validation with SSRF protection
`tapestry-sanitize-filename`	Safe filename generation
`tapestry-safe-download`	Secure file downloads with size limits
`tapestry-vtt-to-text`	VTT subtitle to plain text conversion
`tapestry-extract-html`	Fallback HTML content extractor

Example:

uv run tapestry-validate-url "https://example.com"
uv run tapestry-sanitize-filename "My: Unsafe/Filename"
uv run tapestry-vtt-to-text captions.vtt --output transcript.txt

Dependencies

All Python dependencies are managed via UV and pinned in pyproject.toml:

Package	Purpose
`yt-dlp`	YouTube video/transcript downloads
`trafilatura`	Article extraction
`openai-whisper` (optional)	Transcription for videos without subtitles

Optional system tools (install separately for best results):

reader - Mozilla Readability CLI (npm install -g reader-cli)
pdftotext - PDF text extraction (brew install poppler)

Examples

Example 1: Tapestry Unified Workflow

User: "tapestry https://www.youtube.com/watch?v=dQw4w9WgXcQ"

Claude:
Tapestry Workflow Starting...
Detected: youtube
Extracting YouTube transcript...
Saved transcript: Never Gonna Give You Up.txt

Creating action plan...
Quest: Master Video Production Techniques
Saved plan: Ship-Learn-Next Plan - Master Video Production.md

Tapestry Complete!
Content: Never Gonna Give You Up.txt
Plan: Ship-Learn-Next Plan - Master Video Production.md

Rep 1 (This Week): Film and edit a 60-second video
When will you ship Rep 1?

Example 2: Download YouTube Transcript

User: "Download transcript for https://www.youtube.com/watch?v=VIDEO_ID"

Claude:
Checked available subtitles
Downloaded auto-generated transcript
Converted to readable format
Removed duplicate lines
Saved to: Video Title.txt

Example 3: Extract Article

User: "Extract https://example.com/how-to-build-saas"

Claude:
Using trafilatura for extraction
Extracted article: How to Build a SaaS in 30 Days
Saved to: How to Build a SaaS in 30 Days.txt

Preview (first 10 lines):
[Clean article text without ads or navigation...]

Philosophy

These skills are built on the principle that learning = doing better, not knowing more.

Ship-Learn-Next Framework

Ship: Create something real (code, content, product)
Learn: Honest reflection on what happened
Next: Plan the next iteration based on learnings

100 reps beats 100 hours of study.

Project Structure

tapestry-skills-for-claude-code/
├── plugin.json             # Claude Code plugin manifest
├── pyproject.toml          # Dependencies (UV/pip)
├── src/tapestry/           # Python utilities
│   ├── validate_url.py
│   ├── sanitize_filename.py
│   ├── safe_download.py
│   ├── vtt_to_text.py
│   └── html_extractor.py
├── skills/                 # Claude Code skills
│   ├── tapestry/           # Master skill
│   ├── youtube-transcript/ # YouTube extraction skill
│   ├── article-extractor/  # Article extraction skill
│   └── ship-learn-next/    # Action planning skill
├── tests/                  # Test suite
├── shared/references/      # Security documentation
└── install.sh              # Installation script

Contributing

Found a bug or want to add a feature? Contributions are welcome!

Fork the repository
Create your feature branch (git checkout -b feature/amazing-skill)
Commit your changes (git commit -m 'Add amazing skill')
Push to the branch (git push origin feature/amazing-skill)
Open a Pull Request

License

MIT License - see LICENSE file for details

Acknowledgments

UV: Fast Python package management from Astral
Ship-Learn-Next framework: Inspired by the ShipLearnNext GPT methodology
yt-dlp: Excellent tool for downloading YouTube content
OpenAI Whisper: State-of-the-art speech recognition
Mozilla Readability: Clean article extraction algorithm
trafilatura: Python web scraping and content extraction

Support

Having issues? Please open an issue on GitHub.

Made with Claude Code

Learn more about Claude Code at claude.com/claude-code