# youtube-watcher

> Fetch and read transcripts from YouTube videos. Use when you need to summarize a video, answer questions about its content, or extract information from it.

- Author: albertmusic102
- Repository: dlmmedia/agent-orchestrator
- Version: 20260126105354
- Stars: 0
- Forks: 0
- Last Updated: 2026-02-07
- Source: https://github.com/dlmmedia/agent-orchestrator
- Web: https://mule.run/skillshub/@@dlmmedia/agent-orchestrator~youtube-watcher:20260126105354

---

---
name: youtube-watcher
description: Fetch and read transcripts from YouTube videos. Use when you need to summarize a video, answer questions about its content, or extract information from it.
author: michael gathara
version: 1.0.0
triggers:
  - "watch youtube"
  - "summarize video"
  - "video transcript"
  - "youtube summary"
  - "analyze video"
metadata: {"architect":{"emoji":"📺","requires":{"bins":["yt-dlp"]},"install":[{"id":"brew","kind":"brew","formula":"yt-dlp","bins":["yt-dlp"],"label":"Install yt-dlp (brew)"},{"id":"pip","kind":"pip","package":"yt-dlp","bins":["yt-dlp"],"label":"Install yt-dlp (pip)"}]}}
---

# YouTube Watcher - Architect Agent Skill

Fetch transcripts from YouTube videos to enable summarization, Q&A, and content extraction.

## Installation

### Install yt-dlp

```bash
# macOS (Homebrew)
brew install yt-dlp

# pip (any platform)
pip install yt-dlp

# Or download binary directly
curl -L https://github.com/yt-dlp/yt-dlp/releases/latest/download/yt-dlp -o /usr/local/bin/yt-dlp
chmod a+rx /usr/local/bin/yt-dlp
```

## Usage

### Get Transcript

Retrieve the text transcript of a video:

```bash
python3 {baseDir}/scripts/get_transcript.py "https://www.youtube.com/watch?v=VIDEO_ID"
```

### Output Format

The script outputs the full transcript text with timestamps, making it easy to:
- Search for specific topics
- Generate summaries
- Answer questions about the video content
- Extract quotes and key points

## Example Use Cases

### 1. Summarize a Video

When the user asks "Can you summarize this YouTube video?":

```bash
# Step 1: Get the transcript
python3 {baseDir}/scripts/get_transcript.py "https://www.youtube.com/watch?v=VIDEO_ID"

# Step 2: Read the output and provide a summary
```

**Prompting approach:**
1. Fetch the transcript
2. Identify main topics and key points
3. Provide a concise summary (e.g., 3-5 bullet points for short videos, longer for extensive content)

### 2. Answer Questions About Video Content

When the user asks "What does the speaker say about X?":

```bash
# Fetch transcript
python3 {baseDir}/scripts/get_transcript.py "https://www.youtube.com/watch?v=VIDEO_ID"
```

Then search the transcript for mentions of the topic and provide relevant quotes with context.

### 3. Extract Key Information

For tutorials, lectures, or educational content:
- Extract step-by-step instructions
- List tools or resources mentioned
- Identify timestamps for important sections

### 4. Compare Multiple Videos

When analyzing related content:
1. Fetch transcripts from multiple videos
2. Compare perspectives, information, or approaches
3. Synthesize findings

## Supported Video Types

- Public YouTube videos
- Unlisted videos (with link)
- Videos with:
  - Manual closed captions (CC)
  - Auto-generated subtitles
  - Multiple language tracks

## Troubleshooting

### No Subtitles Available

Some videos don't have captions. The script will return an error message. In this case:
- Inform the user that the video has no available transcript
- Suggest they try a different video or check if the creator has added captions

### Age-Restricted Content

Age-restricted videos may require authentication. Consider using cookies if needed:

```bash
yt-dlp --cookies-from-browser chrome "VIDEO_URL"
```

### Video Not Available

Private videos or region-restricted content cannot be accessed. The script will indicate this in the error message.

## Best Practices

1. **Always verify the URL**: Ensure it's a valid YouTube video link
2. **Check transcript quality**: Auto-generated captions may have errors
3. **Note the video length**: Very long videos will produce large transcripts
4. **Respect copyright**: Use transcripts for analysis, not wholesale reproduction

## Limitations

- Requires `yt-dlp` to be installed and in PATH
- Only works with videos that have subtitles/captions
- Cannot access private or restricted videos without authentication
- Auto-generated captions may contain transcription errors

## Related Tools

- **yt-dlp**: Full-featured YouTube downloader
- **youtube-transcript-api**: Python library for YouTube transcripts
- **Whisper**: For videos without captions (audio transcription)