zhongwei/gh-machu-gwu-sanhe-claude-code-plugins-plugins-social-media-network-youtube
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Initial commit

Browse Source
This commit is contained in:
Zhongwei Li
2025-11-30 08:38:41 +08:00
commit 7d5b628e7d
13 changed files with 1610 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
.claude-plugin/plugin.json Normal file
View File

@@ -0,0 +1,14 @@
{
"name": "youtube",
"description": "Youtube toolkit for Claude Code.",
"version": "1.0.0",
"author": {
"name": "Sanhe Hu"
},
"skills": [
"./skills"
],
"commands": [
"./commands"
]
}

3
README.md Normal file
View File

@@ -0,0 +1,3 @@
# youtube
Youtube toolkit for Claude Code.

121
commands/enrich-citations.md Normal file
View File

@@ -0,0 +1,121 @@
---
name: enrich-citations
description: Add authoritative source links for all facts, citations, and references in markdown content
---
Process markdown content and enrich it with authoritative citation links:
<input_markdown_content>
$ARGUMENTS
</input_markdown_content>
## Input Processing
The command accepts ARGUMENTS which can be either:
1. **Absolute file path**: If the ARGUMENTS starts with `/`, treat it as a file path
- Read the markdown content from the specified file
- Example: `/enrich-citations /Users/sanhehu/Documents/article.md`
2. **Inline content**: If ARGUMENTS is not an absolute path, treat it as inline markdown text
- Process the text directly
- Example: `/enrich-citations This article mentions React and TypeScript frameworks.`
## Workflow
### 1. Load Content
- If absolute path provided: Read content from file
- If inline content provided: Use the provided text directly
### 2. Enrich Citations
Use the `enrich-citations` agent skill to process the markdown content:
**Identify all references:**
- External sources and research papers
- Tools, software, frameworks, libraries
- Products and services
- Organizations and institutions
- Technical concepts and standards
- People and experts
**For each identified reference:**
1. Perform web search to find the authoritative source
2. Verify the URL is current and accessible
3. Replace plain text with markdown hyperlinks: `[Reference](URL)`
4. **Add spaces around hyperlinks**: `text [link](url) text`
**Citation priority:**
- Official sources (project homepages, official documentation)
- Primary sources (original research papers, publications)
- Authoritative organizations (standards bodies, academic institutions)
- Reputable publications (tech blogs, news sites)
- Community resources (GitHub, Stack Overflow)
- General references (Wikipedia as last resort)
**Quality requirements:**
- Verify all links are valid and working
- Ensure proper markdown formatting with spaces
- Preserve all original content (only add hyperlinks)
- Maintain document structure and meaning
### 3. Save Output
Save the enriched document to: `~/tmp/citation_enriched.md`
**Example enrichment:**
Input:
```markdown
# Introduction to Modern JavaScript
JavaScript has evolved significantly with ES6 introducing features like arrow functions.
TypeScript adds static typing to JavaScript and is developed by Microsoft.
```
Output:
```markdown
# Introduction to Modern JavaScript
[JavaScript](https://developer.mozilla.org/en-US/docs/Web/JavaScript) has evolved significantly with [ES6](https://www.ecma-international.org/ecma-262/6.0/) introducing features like arrow functions.
[TypeScript](https://www.typescriptlang.org/) adds static typing to JavaScript and is developed by [Microsoft](https://www.microsoft.com/) .
```
### 4. Display Output Location
After completing the enrichment and saving the document, print the absolute path in a clickable format.
**Format:** `file:///absolute/path/to/citation_enriched.md`
**Example output:**
```
✓ Citation enrichment completed successfully!
✓ Enriched document saved to: file:///Users/sanhehu/tmp/citation_enriched.md
Click the link above to open the document.
```
**Requirements:**
- Use absolute path (not relative path like `~/tmp/...`)
- Include `file://` protocol prefix for clickability
- Provide clear success message
- Make it easy for user to access the enriched document
---
## Usage Examples
**Enrich an existing file:**
```bash
/enrich-citations /Users/sanhehu/Documents/article.md
```
**Enrich inline content:**
```bash
/enrich-citations This article discusses React, Next.js, and Vercel deployment.
```
---
**Note:** This command makes citation enrichment quick and accessible. It automatically performs web searches to find authoritative sources and adds verified hyperlinks to the document.

72
commands/yt-to-md.md Normal file
View File

@@ -0,0 +1,72 @@
---
name: yt-to-md
description: Download, transcribe, and organize YouTube video into structured markdown document
---
Process YouTube video $ARGUMENTS end-to-end and create a well-organized document:
## Workflow
Execute these skills in sequence:
### 1. Download Audio
Use the `youtube-video-to-audio` agent skill to download audio from the provided YouTube URL.
### 2. Transcribe Audio
Use the `transcribe-audio-to-text` agent skill to transcribe the downloaded audio file.
### 3. Clean Up Transcription
Use the `audio-transcription-cleanup` agent skill to transform the raw transcription into a well-structured markdown document.
This step will:
- Detect the primary language and use it for output
- Add semantic paragraph breaks based on topic changes
- Create descriptive section headings
- Fix obvious spelling, terminology, and grammatical errors
- Remove filler words and verbal artifacts
- Maintain all original information from the video
**Output Format:**
```markdown
# [Video Title]
## [Section 1 Heading]
[Cleaned paragraph content...]
[More cleaned content...]
## [Section 2 Heading]
[Content continues...]
## Key Takeaways
- Point 1
- Point 2
- Point 3
```
Save the cleaned document to: `~/tmp/organized_transcript.md`
### 4. Display Output Location
After completing the cleanup and saving the document, print the absolute path of the final organized document in a clickable format.
**Format:** `file:///absolute/path/to/organized_transcript.md`
**Example output:**
```
✓ YouTube video processed successfully!
✓ Organized document saved to: file:///Users/sanhehu/tmp/organized_transcript.md
Click the link above to open the document.
```
**Requirements:**
- Use absolute path (not relative path like `~/tmp/...`)
- Include `file://` protocol prefix for clickability
- Provide clear success message
- Make it easy for user to access the final document
**Note:** This command automates the core workflow from video URL to organized markdown document. Each skill can also be run independently for partial workflows.

81
plugin.lock.json Normal file
View File

@@ -0,0 +1,81 @@
{
"$schema": "internal://schemas/plugin.lock.v1.json",
"pluginId": "gh:MacHu-GWU/sanhe-claude-code-plugins:plugins/social-media-network/youtube",
"normalized": {
"repo": null,
"ref": "refs/tags/v20251128.0",
"commit": "3cdcff17c37ed05dd39655464e9a6341dc088180",
"treeHash": "0df4fe3ca127d4d78db9fa6573360edd6691b523c1c57940ac42eea957b6c88d",
"generatedAt": "2025-11-28T10:12:04.264380Z",
"toolVersion": "publish_plugins.py@0.2.0"
},
"origin": {
"remote": "git@github.com:zhongweili/42plugin-data.git",
"branch": "master",
"commit": "aa1497ed0949fd50e99e70d6324a29c5b34f9390",
"repoRoot": "/Users/zhongweili/projects/openmind/42plugin-data"
},
"manifest": {
"name": "youtube",
"description": "Youtube toolkit for Claude Code.",
"version": "1.0.0"
},
"content": {
"files": [
{
"path": "README.md",
"sha256": "87417563457ce722c372f13f4331678321c1b8193ed81b800984c97a1b1b4f8f"
},
{
"path": ".claude-plugin/plugin.json",
"sha256": "be542de53f0d607ef00f714a666130e8ef99efe83eb435d568de4b68227b79f7"
},
{
"path": "commands/yt-to-md.md",
"sha256": "1b64f6bc6cc2a9b0241f8ffe27790eb599813396ac4acd5be95ce5bc490e002f"
},
{
"path": "commands/enrich-citations.md",
"sha256": "17038f9db9bab9d27d3dbdb2d03b9388081b0f0f6ba4c86f50fbde4ffe93a18c"
},
{
"path": "skills/youtube-video-to-audio/SKILL.md",
"sha256": "90345dafd0294c0fff62e075e2e9bcb86936234c61fb047ec712745daf6b138a"
},
{
"path": "skills/youtube-video-to-audio/scripts/download_audio.py",
"sha256": "5b634eccb702a33ec5b87803de6d20bb03b93086e3926d1875678c029fecf1e0"
},
{
"path": "skills/enrich-citations/SKILL.md",
"sha256": "441142d7f9cf2d239e9f83c17ebca388244908e03d725d8feda0f42b5b10c528"
},
{
"path": "skills/enrich-citations/scripts/enrich_citations.py",
"sha256": "abe901c9480f7f0e7427c870841c89e911719b376bd1b7ac9b92f244befa88b0"
},
{
"path": "skills/audio-transcript-cleanup/SKILL.md",
"sha256": "2d7cbebe9bfd85ff17b72c766a6b93cafca5f38f2c6e51317d761085a54e0dad"
},
{
"path": "skills/audio-transcript-cleanup/scripts/audio_transcript_cleanup.py",
"sha256": "f7f14e690f8cc8c8a279508147a66434cbe1ccd2849c3ec977fec1f14e33f493"
},
{
"path": "skills/transcribe-audio-to-text/SKILL.md",
"sha256": "d7756a805f2ae7ba8544fd608b9c1d8b8761700106654730fa24c9297f60a6c0"
},
{
"path": "skills/transcribe-audio-to-text/scripts/transcribe_audio.py",
"sha256": "e97635f3ae7fb89f4477da6d792cbe190fdce472470eba9ea2b49813f0d502aa"
}
],
"dirSha256": "0df4fe3ca127d4d78db9fa6573360edd6691b523c1c57940ac42eea957b6c88d"
},
"security": {
"scannedAt": null,
"scannerVersion": null,
"flags": []
}
}

54
skills/audio-transcript-cleanup/SKILL.md Normal file
View File

@@ -0,0 +1,54 @@
---
name: audio-transcription-cleanup
description: Transform messy voice transcription text into well-formatted, human-readable documents while preserving original meaning
---
# Audio Transcription Cleanup
Clean up raw audio transcriptions by removing filler words, fixing errors, and adding proper structure.
## Usage
Use the `audio_transcript_cleanup.py` script to process transcript files:
```bash
# Use default output location (~/tmp/cleaned_transcript.md - allows overwrite)
python scripts/audio_transcript_cleanup.py --transcript-file /path/to/transcript.txt
# Specify custom output location (cannot overwrite existing files)
python scripts/audio_transcript_cleanup.py --transcript-file /path/to/transcript.txt --output /path/to/output.md
```
## What It Does
The script automatically:
- Removes verbal artifacts (um, uh, like, you know, 呃, 啊, 那个, etc.)
- Fixes spelling and grammar errors
- Adds semantic paragraph breaks and section headings
- Converts spoken fragments into complete sentences
- Preserves all original information (no summarization)
- Auto-detects language and maintains natural expression
## Options
- `--transcript-file` (required) - Path to the transcript file to clean up
- `--output` (optional) - Custom output path (default: `~/tmp/cleaned_transcript.md`)
## Output Behavior
- **Default location**: `~/tmp/cleaned_transcript.md` - Allows overwrite
- **Custom location**: Cannot overwrite existing files (raises error if file exists)
## Language Support
Auto-detects and works with:
- English
- Chinese (Mandarin, Cantonese)
- Mixed language content
- Multi-speaker transcriptions
## Requirements
- Python 3.11+
- Claude CLI must be installed and accessible
- Transcript file must exist at specified path

329
skills/audio-transcript-cleanup/scripts/audio_transcript_cleanup.py Normal file
View File

@@ -0,0 +1,329 @@
# -*- coding: utf-8 -*-
"""
Audio Transcription Cleanup Script
This module transforms messy voice transcription text into well-formatted, human-readable
markdown documents using Claude CLI. It preserves all original meaning and content while
improving readability through intelligent text cleanup and restructuring.
Key Features:
- Remove verbal artifacts (um, uh, like, filler words in any language)
- Fix spelling and grammar errors
- Add semantic paragraph breaks and section headings
- Support for single-speaker and multi-speaker content
- Multi-language support (English, Chinese, and more)
- Preserve all original information (no summarization)
- Default output location with overwrite capability
- Custom output locations with overwrite protection
Default Behavior:
- Output: ~/tmp/cleaned_transcript.md (allows overwrite)
- Language: Auto-detected from source
- Format: Well-structured markdown with headings
Custom Output:
- Cannot overwrite existing files (raises FileExistsError)
- Creates parent directories automatically
Text Cleanup Operations:
- Remove filler words: um, uh, like, you know, 呃, 啊, 那个, 就是说
- Fix obvious noun errors and typos
- Correct grammar mistakes
- Convert spoken fragments to complete sentences
- Add descriptive section headings
- Organize content into semantic paragraphs
Example Usage:
# Use default output location (allows overwrite)
$ python audio_transcription_cleanup.py --transcript-file /path/to/transcript.txt
# Specify custom output location (cannot overwrite existing file)
$ python audio_transcription_cleanup.py --transcript-file /path/to/transcript.txt --output ~/Documents/cleaned.md
Requirements:
- Claude CLI must be installed and accessible
- Transcript file must exist at specified path
Author: sanhe
Plugin: youtube@sanhe-claude-code-plugins
"""
import subprocess
import argparse
from pathlib import Path
dir_home = Path.home()
path_cleaned_transcript = dir_home / "tmp" / "cleaned_transcript.md"
prompt = """
## Task
Transform the messy voice transcription text provided below into a well-formatted, human-readable document while preserving ALL original meaning and content.
## Key Principles
- **Preserve original meaning**: Do not summarize or omit information
- **Fix transcription artifacts**: Remove filler words, false starts, and repetitions
- **Improve readability**: Organize into proper paragraphs with clear structure
- **Handle multi-speaker content**: Clearly attribute dialogue when multiple speakers are present
- **Enhance document structure**: Add semantic paragraphing and descriptive headings
## Processing Instructions
### Step 1: Analyze the Input
Examine the transcription to identify:
- **Primary language**: Determine the dominant language of the transcription
- **Number of speakers**: Single vs. multi-speaker content
- **Main topics or themes**: Identify distinct topics for sectioning
- **Transcription quality issues**: Filler words, repetitions, false starts, obvious errors
### Step 2: Text Cleanup Operations
1. **Remove verbal artifacts**:
- Filler words: "um", "uh", "like", "you know", "呃", "啊", "那个", "就是说", "嗯", "然后"
- False starts and self-corrections
- Unnecessary repetitions
2. **Fix errors and improve accuracy**:
- **Correct obvious noun errors**: Fix misrecognized names, places, technical terms
- **Fix spelling mistakes**: Correct typos and transcription errors
- **Fix grammar errors**: Subject-verb agreement, tense consistency, word order
- **Clarify ambiguous terms**: Use context to determine correct words when transcription is unclear
3. **Organize content structure**:
- **Semantic paragraphing**: Group related ideas into logical paragraphs based on meaning
- **Add section headings**: Create descriptive headings that summarize each section's content
- Convert spoken fragments into complete sentences
- Create logical flow between sections
4. **For multi-speaker content**:
- Use clear speaker labels (Speaker A, B, or actual names if identified)
- Format as dialogue or meeting notes
- Preserve conversational context
### Step 3: Language and Formatting
**Language Selection Rules** (in priority order):
1. If user explicitly specifies output language → use that language
2. Otherwise → use the primary language of the original transcription
3. Preserve technical terms and proper nouns in their original language/form
**For single-speaker content**:
```markdown
# [Main Topic Title]
## [Section 1 Heading]
[Cleaned paragraph 1 content...]
[Cleaned paragraph 2 content...]
## [Section 2 Heading]
[Cleaned content...]
```
**For multi-speaker content** - Option 1 (Dialogue format):
```markdown
# [Meeting/Discussion Title]
## [Topic 1]
**Speaker A:** [cleaned content]
**Speaker B:** [cleaned content]
## [Topic 2]
**Speaker A:** [cleaned content]
```
**For multi-speaker content** - Option 2 (Meeting notes format):
```markdown
# [Meeting Title]
## [Discussion Topic 1]
Speaker A mentioned that [cleaned content]...
Speaker B responded by explaining [cleaned content]...
## [Discussion Topic 2]
The group discussed [cleaned content]...
```
## Important Constraints
- **NEVER summarize**: Include all information from original transcription
- **NEVER add new substantive information**: Only reorganize and clarify existing content
- **NEVER change the speaker's meaning or intent**
- **DO add section headings**: These are structural additions that help readability
- **Preserve technical terms and specific names** exactly as intended (fix transcription errors only)
## Heading Guidelines
- Section headings should be **concise and descriptive** (3-8 words typically)
- Headings should reflect the **actual content** of each section
- Use appropriate heading levels (H1 for main title, H2 for major sections, H3 for subsections)
- Headings are added to **improve navigation**, not to interpret or editorialize
## Language Support
This skill works with transcriptions in any language:
- **Chinese** (Mandarin, Cantonese, etc.)
- **English**
- **Mixed language** content (preserve code-switching naturally)
- **Other languages** (apply language-appropriate grammar rules)
## OUTPUT INSTRUCTIONS
**CRITICAL**: Your response must contain ONLY the cleaned markdown document.
**DO NOT include**:
- Any explanations before the document
- Any explanations after the document
- Any meta-commentary about the cleaning process
- Any acknowledgments like "Here is the cleaned version"
- Any markdown code fences (no ```markdown blocks)
- Any introductory or concluding remarks
**START your response immediately with the H1 title** (# [Title]) and **END immediately after the last content paragraph**.
Your entire response = the cleaned document itself, nothing more.
## INPUT TRANSCRIPTION:
""".strip()
def cleanup_transcript(path_transcript: Path):
"""
Clean up audio transcription and save to file.
Args:
path_transcript: Path to the raw transcript file
"""
# Read transcript content
content = path_transcript.read_text(encoding="utf-8")
# Call Claude CLI to process the transcript
args = [
"claude",
"--append-system-prompt",
prompt,
"--print",
content,
]
result = subprocess.run(
args,
capture_output=True,
text=True,
)
if result.returncode != 0:
raise RuntimeError(f"Claude CLI failed: {result.stderr}")
# Save the cleaned output
cleaned_transcript = result.stdout
return cleaned_transcript
def main():
"""
Main CLI entry point for cleaning up audio transcriptions.
This script transforms messy voice transcription text into well-formatted,
human-readable markdown documents while preserving the original meaning.
Example usage:
python audio_transcription_cleanup.py --transcript-file "/path/to/transcript.txt"
What it does:
- Removes filler words and verbal artifacts (um, uh, like, etc.)
- Fixes obvious spelling and grammar errors
- Adds semantic paragraph breaks and section headings
- Preserves all original information (no summarization)
- Maintains the speaker's meaning and intent
Requirements:
- Transcript file must exist
- Claude CLI must be installed and accessible
"""
parser = argparse.ArgumentParser(
description="Clean up audio transcriptions into well-formatted markdown documents",
formatter_class=argparse.RawDescriptionHelpFormatter,
epilog="""
Examples:
# Use default output location (allows overwrite)
%(prog)s --transcript-file "~/tmp/transcript.txt"
# Specify custom output location (cannot overwrite existing file)
%(prog)s --transcript-file "~/tmp/transcript.txt" --output "~/Documents/cleaned.md"
Cleanup Operations:
- Remove verbal artifacts: um, uh, like, you know, 呃, 啊, 那个
- Fix spelling and grammar errors
- Add semantic paragraphs and section headings
- Convert spoken fragments into complete sentences
- Preserve all original information
Output Behavior:
- Default location (~/tmp/cleaned_transcript.md): Allows overwrite
- Custom location: Cannot overwrite existing files (will raise error)
Note:
This script uses Claude CLI to perform intelligent transcript cleanup.
All original information is preserved - no summarization occurs.
""",
)
parser.add_argument(
"--transcript-file",
type=str,
required=True,
help=f"Path to the transcript file to clean up (required)",
)
parser.add_argument(
"--output",
type=str,
default=None,
help=f"Output file path for cleaned transcript (default: {path_cleaned_transcript}). Note: Default location allows overwrite, custom locations cannot overwrite existing files.",
)
args = parser.parse_args()
# Convert to Path objects and expand user home directory
path_transcript = Path(args.transcript_file).expanduser()
# Determine output path
if args.output:
path_output = Path(args.output).expanduser()
# For custom locations, check if file already exists
if (path_output != path_cleaned_transcript) and path_output.exists():
raise FileExistsError(
f"Output file already exists at {path_output}. "
f"Please choose a different location or remove the existing file. "
f"(Default location {path_cleaned_transcript} allows overwrite)"
)
else:
# Use default location (allows overwrite)
path_output = path_cleaned_transcript
# Ensure output directory exists
path_output.parent.mkdir(parents=True, exist_ok=True)
# Clean up the transcript
cleaned_transcript = cleanup_transcript(
path_transcript=path_transcript,
)
# Save to output file
path_output.write_text(cleaned_transcript, encoding="utf-8")
# Print success message with clickable file path
absolute_path = path_output.resolve()
print(f"✓ Transcript cleanup completed successfully!")
print(f"✓ Cleaned transcript saved to: file://{absolute_path}")
print(f"\nClick the link above to open the document.")
if __name__ == "__main__":
main()

56
skills/enrich-citations/SKILL.md Normal file
View File

@@ -0,0 +1,56 @@
---
name: enrich-citations
description: Find and add authoritative source links for all facts, citations, and references in markdown documents
---
# Enrich Citations
Enhance markdown documents by finding and adding authoritative source links for mentioned facts, tools, products, research, and references.
## Usage
Use the `enrich_citations.py` script to process markdown documents:
```bash
# Use default output location (~/tmp/citation_enriched.md - allows overwrite)
python scripts/enrich_citations.py --document-file /path/to/document.md
# Specify custom output location (cannot overwrite existing files)
python scripts/enrich_citations.py --document-file /path/to/document.md --output /path/to/output.md
```
## What It Does
The script automatically:
- Identifies all references (tools, research, products, organizations, people, standards)
- Performs web search to find authoritative sources
- Adds markdown hyperlinks with proper spacing: `[Reference](URL)`
- Verifies all URLs are valid and accessible
- Preserves all original content (only adds hyperlinks, no text changes)
- Prioritizes official sources and documentation
## Options
- `--document-file` (required) - Path to the markdown document to enrich
- `--output` (optional) - Custom output path (default: `~/tmp/citation_enriched.md`)
## Output Behavior
- **Default location**: `~/tmp/citation_enriched.md` - Allows overwrite
- **Custom location**: Cannot overwrite existing files (raises error if file exists)
## Reference Types Identified
- External sources and research papers
- Tools, software, frameworks, libraries
- Products and services
- Organizations and institutions
- Technical concepts and standards (RFC, W3C, APIs)
- People and experts
## Requirements
- Python 3.11+
- Claude CLI must be installed and accessible
- Internet connection (for web searches)
- Document file must exist at specified path

323
skills/enrich-citations/scripts/enrich_citations.py Normal file
View File

@@ -0,0 +1,323 @@
# -*- coding: utf-8 -*-
"""
Example usage::
# Use default output location (allows overwrite)
python /Users/sanhehu/Documents/GitHub/sanhe-claude-code-plugins/plugins/social-media-network/youtube/skills/enrich-citations/scripts/enrich_citations.py --document-file /Users/sanhehu/tmp/cleaned_transcript.md
# Specify custom output location (cannot overwrite existing file)
python /Users/sanhehu/Documents/GitHub/sanhe-claude-code-plugins/plugins/social-media-network/youtube/skills/enrich-citations/scripts/enrich_citations.py --document-file /Users/sanhehu/tmp/cleaned_transcript.md --output ~/Documents/enriched_document.md
"""
import subprocess
import argparse
from pathlib import Path
dir_home = Path.home()
path_enriched_document = dir_home / "tmp" / "citation_enriched.md"
prompt = """
## Task
Enhance the markdown document provided below by finding and adding authoritative source links for all mentioned facts, citations, tools, products, research, and references.
## Key Principles
- **Find authoritative sources**: Search for the original, most credible source for each reference
- **Verify accuracy**: Ensure links point to the correct, current information
- **Preserve content**: Only add hyperlinks, don't change the document's text or meaning
- **Proper formatting**: Use markdown hyperlink syntax with spaces for proper rendering
## Processing Instructions
### Step 1: Identify References
Read through the entire markdown document and identify all mentions of:
1. **External sources and research**:
- Academic papers, studies, reports
- Books, articles, blog posts
- Research findings or statistics
2. **Tools and software**:
- Applications, frameworks, libraries
- Programming languages, platforms
- Development tools, services
3. **Products and services**:
- Commercial products
- SaaS platforms
- Hardware or software products
4. **Organizations and institutions**:
- Companies, universities
- Standards bodies, foundations
- Government agencies
5. **Technical concepts and standards**:
- Specifications (RFC, W3C, etc.)
- APIs, protocols
- Industry standards
6. **People and experts**:
- Authors, researchers
- Industry leaders
- Subject matter experts
### Step 2: Web Search for Sources
For **each identified reference**:
1. **Search for the authoritative source**:
- Use web search to find the official or most credible source
- Prioritize: official websites > documentation > reputable publications
- For academic references: search for DOI, arXiv, official publication
2. **Verify the source**:
- Ensure the URL is current and accessible
- Check that content matches the reference
- Confirm it's the most authoritative source available
3. **Select the best link**:
- Official documentation or homepage for tools/products
- Original publication for research/articles
- Wikipedia for general concepts (if no better source exists)
- GitHub repository for open source projects
### Step 3: Add Hyperlinks
Replace plain text references with markdown hyperlinks:
**Format:** `[Reference Text](URL)`
**Critical formatting requirements:**
- **Add spaces around hyperlinks**: `text [link](url) text` (not `text[link](url)text`)
- Use descriptive link text (the actual reference name, not "click here")
- Maintain the original sentence structure
- Preserve all other content unchanged
**Examples:**
Before:
```
According to the 2023 Stack Overflow Survey, JavaScript is the most popular language.
```
After:
```
According to the [2023 Stack Overflow Survey](https://survey.stackoverflow.co/2023/) , JavaScript is the most popular language.
```
Before (Chinese):
```
文章提到了GPT-4的能力提升
```
After (Chinese):
```
文章提到了 [GPT-4](https://openai.com/gpt-4) 的能力提升
```
### Step 4: Quality Assurance
Before finalizing:
1. **Verify all links**:
- Test that URLs are valid and accessible
- Check for broken links or redirects
- Ensure HTTPS when available
2. **Check formatting**:
- Confirm spaces around hyperlinks
- Verify markdown syntax is correct
- Ensure links render properly
3. **Review accuracy**:
- Links point to correct resources
- No duplicate or conflicting sources
- All identified references are addressed
## Important Constraints
- **NEVER change the document content**: Only add hyperlinks to existing text
- **NEVER add new information**: Don't insert citations not mentioned in the original
- **NEVER remove existing content**: Preserve all original text
- **NEVER summarize or paraphrase**: Keep exact wording
- **DO add spaces around hyperlinks**: Essential for proper rendering
- **DO verify all sources**: Ensure accuracy and accessibility
## Source Priority Guidelines
When multiple sources exist, prioritize in this order:
1. **Official sources**: Project homepages, official documentation
2. **Primary sources**: Original research papers, first publications
3. **Authoritative organizations**: Standards bodies, academic institutions
4. **Reputable publications**: Well-known tech blogs, news sites
5. **Community resources**: GitHub, Stack Overflow (for code/tools)
6. **General references**: Wikipedia, encyclopedias (as last resort)
## Language Support
This skill works with documents in any language:
- Respect the document's primary language
- Use appropriate sources for the language/region
- Maintain original text while adding hyperlinks
- For multilingual documents, find sources matching each language section when appropriate
## OUTPUT INSTRUCTIONS
**CRITICAL**: Your response must contain ONLY the enhanced markdown document with citations added.
**DO NOT include**:
- Any explanations before the document
- Any explanations after the document
- Any meta-commentary about the enrichment process
- Any acknowledgments like "Here is the enriched version"
- Any markdown code fences (no ```markdown blocks)
- Any introductory or concluding remarks
**START your response immediately with the first line of the document** and **END immediately after the last line**.
Your entire response = the enriched document itself, nothing more.
## INPUT DOCUMENT:
""".strip()
def enrich_citations(path_document: Path):
"""
Enrich markdown document with authoritative citation links.
Args:
path_document: Path to the markdown document file
"""
# Read document content
content = path_document.read_text(encoding="utf-8")
# Call Claude CLI to process the document
args = [
"claude",
"--append-system-prompt",
prompt,
"--print",
content,
]
result = subprocess.run(
args,
capture_output=True,
text=True,
)
if result.returncode != 0:
raise RuntimeError(f"Claude CLI failed: {result.stderr}")
# Return the enriched document
enriched_document = result.stdout
return enriched_document
def main():
"""
Main CLI entry point for enriching citations in markdown documents.
This script enhances markdown documents by finding and adding authoritative
source links for all mentioned facts, citations, and references.
Example usage:
python enrich_citations.py --document-file "/path/to/document.md"
What it does:
- Identifies all references (tools, research, products, people, etc.)
- Performs web search to find authoritative sources
- Adds markdown hyperlinks with proper spacing
- Verifies all URLs are valid and current
- Preserves all original content (only adds links)
Requirements:
- Document file must exist
- Claude CLI must be installed and accessible
- Internet connection required for web searches
"""
parser = argparse.ArgumentParser(
description="Enrich markdown documents with authoritative citation links",
formatter_class=argparse.RawDescriptionHelpFormatter,
epilog="""
Examples:
# Use default output location (allows overwrite)
%(prog)s --document-file "~/tmp/cleaned_transcript.md"
# Specify custom output location (cannot overwrite existing file)
%(prog)s --document-file "~/tmp/cleaned_transcript.md" --output "~/Documents/enriched.md"
Citation Enrichment:
- Identifies facts, tools, products, research, and references
- Web search to find authoritative sources
- Adds markdown hyperlinks: [Title](URL)
- Verifies all links are valid and current
- Maintains proper spacing around hyperlinks
- Preserves original document content
Output Behavior:
- Default location (~/tmp/citation_enriched.md): Allows overwrite
- Custom location: Cannot overwrite existing files (will raise error)
Note:
This script uses Claude CLI with web search to perform intelligent citation enrichment.
All original content is preserved - only hyperlinks are added.
""",
)
parser.add_argument(
"--document-file",
type=str,
required=True,
help=f"Path to the markdown document to enrich (required)",
)
parser.add_argument(
"--output",
type=str,
default=None,
help=f"Output file path for enriched document (default: {path_enriched_document}). Note: Default location allows overwrite, custom locations cannot overwrite existing files.",
)
args = parser.parse_args()
# Convert to Path objects and expand user home directory
path_document = Path(args.document_file).expanduser()
# Determine output path
if args.output:
path_output = Path(args.output).expanduser()
# For custom locations, check if file already exists
if (path_output != path_enriched_document) and path_output.exists():
raise FileExistsError(
f"Output file already exists at {path_output}. "
f"Please choose a different location or remove the existing file. "
f"(Default location {path_enriched_document} allows overwrite)"
)
else:
# Use default location (allows overwrite)
path_output = path_enriched_document
# Ensure output directory exists
path_output.parent.mkdir(parents=True, exist_ok=True)
# Enrich citations in the document
enriched_document = enrich_citations(
path_document=path_document,
)
# Save to output file
path_output.write_text(enriched_document, encoding="utf-8")
# Print success message with clickable file path
absolute_path = path_output.resolve()
print(f"✓ Citation enrichment completed successfully!")
print(f"✓ Enriched document saved to: file://{absolute_path}")
print(f"\nClick the link above to open the document.")
if __name__ == "__main__":
main()

45
skills/transcribe-audio-to-text/SKILL.md Normal file
View File

@@ -0,0 +1,45 @@
---
name: transcribe-audio-to-text
description: Transcribe audio files to text using audinota cli
---
# Transcribe Audio to Text
Convert audio files to text using audinota transcription service.
## Usage
```bash
# Use default audio file (output from youtube-video-to-audio skill)
python scripts/transcribe_audio.py
# Specify custom audio file
python scripts/transcribe_audio.py --audio-file-path "/path/to/audio.mp3"
```
## Options
- `--audio-file-path` - Path to audio file to transcribe (default: ~/tmp/download_audio_result.mp3)
## Examples
```bash
# Transcribe default audio file
python scripts/transcribe_audio.py
# Transcribe custom audio file
python scripts/transcribe_audio.py --audio-file-path "~/Music/podcast.mp3"
# Get help
python scripts/transcribe_audio.py -h
```
## Requirements
- Python 3.11+
- audinota installed at `~/Documents/GitHub/audinota-project/.venv/bin/audinota`
- Audio file must exist at specified path
## Integration
This skill works seamlessly with the `youtube-video-to-audio` skill. By default, it transcribes the audio file downloaded by that skill at `~/tmp/download_audio_result.mp3`.

181
skills/transcribe-audio-to-text/scripts/transcribe_audio.py Normal file
View File

@@ -0,0 +1,181 @@
# -*- coding: utf-8 -*-
"""
Audio Transcription Script
This module provides functionality to transcribe audio files to text using the audinota CLI.
It supports both default and custom output locations, with built-in protection against
accidentally overwriting existing files at custom locations.
Key Features:
- Transcribe audio files using audinota
- Default output location with overwrite capability
- Custom output locations with overwrite protection
- Automatic directory creation for custom paths
- Integration with youtube-video-to-audio skill workflow
Default Behavior:
- Input: ~/tmp/download_audio_result.mp3
- Output: ~/tmp/download_audio_result.txt (allows overwrite)
Custom Output:
- Cannot overwrite existing files (raises FileExistsError)
- Creates parent directories automatically
Example Usage:
# Use defaults (allows overwrite)
$ python transcribe_audio.py
# Custom audio file
$ python transcribe_audio.py --audio-file-path "/path/to/audio.mp3"
# Custom output (cannot overwrite existing)
$ python transcribe_audio.py --audio-file-path "/path/to/audio.mp3" --output "~/Documents/transcript.txt"
Requirements:
- audinota installed at ~/Documents/GitHub/audinota-project/.venv/bin/audinota
- Audio file must exist at specified path
Author: sanhe
Plugin: youtube@sanhe-claude-code-plugins
"""
import subprocess
import argparse
from pathlib import Path
dir_home = Path.home()
default_audio_path = dir_home / "tmp" / "download_audio_result.mp3"
default_transcript_path = dir_home / "tmp" / "download_audio_result.txt"
def transcribe_audio(path_audio: Path, path_output: Path = None):
"""
Transcribe an audio file to text using audinota.
Args:
path_audio: Path to the audio file to transcribe
path_output: Path to save the transcript (optional, audinota will use default if not specified)
Raises:
subprocess.CalledProcessError: If audinota transcription fails
"""
path_audinota = (
dir_home
/ "Documents"
/ "GitHub"
/ "audinota-project"
/ ".venv"
/ "bin"
/ "audinota"
)
if not path_audinota.exists():
raise FileNotFoundError(f"audinota not found at {path_audinota}")
if not path_audio.exists():
raise FileNotFoundError(f"Audio file not found at {path_audio}")
args = [
str(path_audinota),
"transcribe",
"--input",
str(path_audio),
]
# Add output argument if specified
if path_output:
args.extend(["--output", str(path_output)])
print(f"Transcribing audio file: {path_audio}")
subprocess.run(args, check=True)
output_location = path_output if path_output else "default location"
print(f"✓ Transcription completed successfully")
print(f"✓ Saved to: {output_location}")
def main():
"""
Main CLI entry point for transcribing audio files to text.
This script uses audinota to transcribe audio files to text.
By default, it transcribes the audio file downloaded by the youtube-video-to-audio skill.
Example usage:
# Use default paths (allows overwrite)
python transcribe_audio.py
# Specify custom audio file with default output
python transcribe_audio.py --audio-file-path "/path/to/audio.mp3"
# Specify both custom audio and output (cannot overwrite existing file)
python transcribe_audio.py --audio-file-path "/path/to/audio.mp3" --output "~/Documents/transcript.txt"
Requirements:
- audinota must be installed at ~/Documents/GitHub/audinota-project/.venv/bin/audinota
- Audio file must exist at the specified path
Note:
- Default location (~/tmp/download_audio_result.txt) allows overwrite
- Custom output locations cannot overwrite existing files
"""
parser = argparse.ArgumentParser(
description="Transcribe audio files to text using audinota",
formatter_class=argparse.RawDescriptionHelpFormatter,
epilog="""
Examples:
%(prog)s
%(prog)s --audio-file-path "/path/to/audio.mp3"
%(prog)s --audio-file-path "~/Music/podcast.mp3" --output "~/Documents/transcript.txt"
Notes:
- The default audio file path is the output of youtube-video-to-audio skill
- Default output location (~/tmp/download_audio_result.txt) allows overwrite
- Custom output locations cannot overwrite existing files
- Requires audinota to be installed at ~/Documents/GitHub/audinota-project/.venv/bin/audinota
""",
)
parser.add_argument(
"--audio-file-path",
type=str,
default=str(default_audio_path),
help=f"Path to the audio file to transcribe (default: {default_audio_path})",
)
parser.add_argument(
"--output",
type=str,
default=None,
help=f"Output file path for transcript (default: {default_transcript_path}). Note: Default location allows overwrite, custom locations cannot overwrite existing files.",
)
args = parser.parse_args()
# Convert to Path object and expand user home directory
audio_path = Path(args.audio_file_path).expanduser()
# Determine output path and check for existing files
if args.output:
path_output = Path(args.output).expanduser()
# For custom locations, check if file already exists
if (path_output != default_transcript_path) and (path_output.exists()):
raise FileExistsError(
f"Output file already exists at {path_output}. "
f"Please choose a different location or remove the existing file. "
f"(Default location {default_transcript_path} allows overwrite)"
)
# Ensure output directory exists
path_output.parent.mkdir(parents=True, exist_ok=True)
else:
# Use default location (allows overwrite)
path_output = default_transcript_path
path_output.unlink(missing_ok=True)
# Transcribe the audio
transcribe_audio(path_audio=audio_path, path_output=path_output)
if __name__ == "__main__":
main()

41
skills/youtube-video-to-audio/SKILL.md Normal file
View File

@@ -0,0 +1,41 @@
---
name: youtube-video-to-audio
description: Download YouTube videos as audio files using yt-dlp
---
# YouTube Video to Audio
Extract and download audio from YouTube videos.
## Usage
```bash
python scripts/download_audio.py --video-url "https://www.youtube.com/watch?v=VIDEO_ID"
```
## Options
- `--video-url` (required) - YouTube video URL to download
- `--quality` - Quality selector (default: "bestaudio[abr<=64]/worstaudio")
- `--audio-format` - Output format: mp3, aac, wav (default: mp3)
- `--bitrate` - Audio bitrate: 64K, 128K, 192K, 320K (default: 64K)
- `--output` - Custom output path (default: ~/tmp/download_audio_result.mp3)
## Examples
```bash
# Basic usage (default: 64K bitrate MP3 to ~/tmp/download_audio_result.mp3)
python scripts/download_audio.py --video-url "https://www.youtube.com/watch?v=d6rZtgHcbWA"
# High quality with custom bitrate
python scripts/download_audio.py --video-url "https://youtu.be/xyz" --quality "bestaudio" --bitrate "128K"
# Custom output location
python scripts/download_audio.py --video-url "https://youtu.be/xyz" --output "/path/to/output.mp3"
```
## Requirements
- Python 3.11+
- ffmpeg installed at `~/ffmpeg`
- yt-dlp (auto-downloaded on first run)

290
skills/youtube-video-to-audio/scripts/download_audio.py Normal file
View File

@@ -0,0 +1,290 @@
# -*- coding: utf-8 -*-
"""
YouTube Audio Downloader
This module provides functionality to download audio from YouTube videos using yt-dlp.
It supports cross-platform operation (Windows, macOS, Linux) and includes automatic
yt-dlp binary management with built-in protection against overwriting existing files.
Key Features:
- Download audio from YouTube videos as MP3 files
- Automatic yt-dlp binary download and updates
- Cross-platform support (Windows, macOS, Linux)
- Customizable audio quality and bitrate
- Default output location with overwrite capability
- Custom output locations with overwrite protection
- Automatic directory creation for custom paths
Default Behavior:
- Output: ~/tmp/download_audio_result.mp3 (allows overwrite)
- Quality: bestaudio[abr<=64]/worstaudio
- Format: MP3
- Bitrate: 64K
Custom Output:
- Cannot overwrite existing files (raises FileExistsError)
- Creates parent directories automatically
Example Usage:
# Download with defaults (allows overwrite)
$ python download_audio.py --video-url "https://www.youtube.com/watch?v=d6rZtgHcbWA"
# Custom quality and bitrate
$ python download_audio.py --video-url "https://youtu.be/xyz" --quality "bestaudio" --bitrate "128K"
# Custom output location (cannot overwrite existing)
$ python download_audio.py --video-url "https://youtu.be/xyz" --output "/path/to/audio.mp3"
Requirements:
- ffmpeg installed at ~/ffmpeg
- Internet connection for downloading yt-dlp and videos
- yt-dlp will be automatically downloaded if not present
Platform Support:
- Windows: Downloads yt-dlp.exe
- macOS: Downloads yt-dlp_macos
- Linux: Downloads yt-dlp_linux
Author: sanhe
Plugin: youtube@sanhe-claude-code-plugins
"""
import sys
import os
import json
import subprocess
import urllib.request
import argparse
from pathlib import Path
# Detect OS
IS_WINDOWS = False
IS_MACOS = False
IS_LINUX = False
_platform = sys.platform
if _platform in ["win32", "cygwin"]:
IS_WINDOWS = True
elif _platform == "darwin":
IS_MACOS = True
elif _platform == "linux":
IS_LINUX = True
else:
raise RuntimeError(f"Unsupported platform: {_platform}")
dir_home = Path.home()
dir_tmp = dir_home / "tmp"
try:
dir_tmp.mkdir(exist_ok=True)
except Exception: # pragma: no cover
pass
path_audio = dir_tmp / "download_audio_result.mp3"
if IS_WINDOWS:
filename = "yt-dlp.exe"
path_yt_dlp = dir_home / "yt-dlp.exe"
elif IS_MACOS:
filename = "yt-dlp_macos"
path_yt_dlp = dir_home / "yt-dlp_macos"
elif IS_LINUX:
filename = "yt-dlp_linux"
path_yt_dlp = dir_home / "yt-dlp_linux"
else:
raise RuntimeError("Unsupported operating system")
path_ffmpeg = dir_home / "ffmpeg"
def get_latest_yt_dlp_release() -> str:
"""Get the latest yt-dlp release version from GitHub API."""
api_url = "https://api.github.com/repos/yt-dlp/yt-dlp/releases/latest"
with urllib.request.urlopen(api_url) as response:
data = json.loads(response.read().decode())
tag_name = data["tag_name"]
return tag_name
def download_yt_dlp():
"""Download the appropriate yt-dlp binary based on the OS."""
# Get latest release version
latest_version = get_latest_yt_dlp_release()
print(f"Latest yt-dlp version: {latest_version}")
# Determine download URL and local path based on OS
download_url = f"https://github.com/yt-dlp/yt-dlp/releases/download/{latest_version}/{filename}"
print(f"Downloading {filename} from {download_url}...")
print(f"Saving to {path_yt_dlp}...")
# Download the file
urllib.request.urlretrieve(download_url, path_yt_dlp)
# Set executable permissions for macOS and Linux
if IS_MACOS or IS_LINUX:
os.chmod(path_yt_dlp, 0o755)
print(f"Successfully downloaded yt-dlp to {path_yt_dlp}")
def download_audio(
video_url: str,
quality: str = "bestaudio[abr<=64]/worstaudio",
audio_format: str = "mp3",
bitrate: str = "64K",
output_path: str = None,
):
"""
Download audio from a YouTube video using yt-dlp.
Args:
video_url: YouTube video URL to download
quality: yt-dlp quality selector (default: "bestaudio[abr<=64]/worstaudio")
audio_format: Output audio format (default: "mp3")
bitrate: Audio bitrate (default: "64K")
output_path: Custom output path (default: ~/tmp/download_audio_result.mp3)
Returns:
Path to the downloaded audio file
"""
output = output_path if output_path else str(path_audio)
args = [
f"{path_yt_dlp}",
"-f",
quality,
"--extract-audio",
"--audio-format",
audio_format,
"--audio-quality",
bitrate,
"-o",
output,
"--restrict-filenames",
"--ffmpeg-location",
f"{str(path_ffmpeg)}",
video_url,
]
result = subprocess.run(args, check=True, capture_output=True)
return output
def main():
"""
Main CLI entry point for downloading YouTube audio.
This script downloads audio from YouTube videos using yt-dlp and ffmpeg.
It automatically downloads yt-dlp if not present in the home directory.
Example usage:
# Use default output location (allows overwrite)
python download_audio.py --video-url "https://www.youtube.com/watch?v=d6rZtgHcbWA"
# Specify custom quality and bitrate
python download_audio.py --video-url "https://youtu.be/xyz" --quality "bestaudio" --bitrate "128K"
# Specify custom output location (cannot overwrite existing file)
python download_audio.py --video-url "https://youtu.be/xyz" --output "/path/to/output.mp3"
Requirements:
- ffmpeg must be installed at ~/ffmpeg
- yt-dlp will be automatically downloaded to home directory if not present
Note:
- Default location (~/tmp/download_audio_result.mp3) allows overwrite
- Custom output locations cannot overwrite existing files
"""
parser = argparse.ArgumentParser(
description="Download audio from YouTube videos as MP3 files using yt-dlp",
formatter_class=argparse.RawDescriptionHelpFormatter,
epilog="""
Examples:
%(prog)s --video-url "https://www.youtube.com/watch?v=d6rZtgHcbWA"
%(prog)s --video-url "https://youtu.be/xyz" --quality "bestaudio" --bitrate "128K"
%(prog)s --video-url "https://youtu.be/xyz" --output "/custom/path/audio.mp3"
Quality options:
bestaudio - Best quality audio available
bestaudio[abr<=64] - Best audio with max 64kbps bitrate
worstaudio - Lowest quality audio (smallest file)
""",
)
# Required arguments
parser.add_argument(
"--video-url",
type=str,
required=True,
help="YouTube video URL to download (required)",
)
# Optional arguments for yt-dlp customization
parser.add_argument(
"--quality",
type=str,
default="bestaudio[abr<=64]/worstaudio",
help='Quality selector for yt-dlp (default: "bestaudio[abr<=64]/worstaudio")',
)
parser.add_argument(
"--audio-format",
type=str,
default="mp3",
help="Output audio format: mp3, aac, wav, etc. (default: mp3)",
)
parser.add_argument(
"--bitrate",
type=str,
default="64K",
help="Audio bitrate: 64K, 128K, 192K, 320K, etc. (default: 64K)",
)
parser.add_argument(
"--output",
type=str,
default=None,
help=f"Custom output file path (default: {path_audio}). Note: Default location allows overwrite, custom locations cannot overwrite existing files.",
)
args = parser.parse_args()
# Ensure yt-dlp is downloaded
if not path_yt_dlp.exists():
print("yt-dlp not found, downloading...")
download_yt_dlp()
# Determine output path and check for existing files
if args.output:
path_output = Path(args.output).expanduser()
# For custom locations, check if file already exists
if (path_output != path_audio) and path_output.exists():
raise FileExistsError(
f"Output file already exists at {path_output}. "
f"Please choose a different location or remove the existing file. "
f"(Default location {path_audio} allows overwrite)"
)
# Ensure output directory exists
path_output.parent.mkdir(parents=True, exist_ok=True)
output_path_str = str(path_output)
else:
# Use default location (allows overwrite)
path_audio.unlink(missing_ok=True)
output_path_str = None
# Download the audio
output_file = download_audio(
video_url=args.video_url,
quality=args.quality,
audio_format=args.audio_format,
bitrate=args.bitrate,
output_path=output_path_str,
)
print(f"✓ Audio successfully downloaded from {args.video_url}")
print(f"✓ Saved to: {output_file}")
if __name__ == "__main__":
main()