← Back to Skills Marketplace
michaelgathara

YouTube Watcher

by Michael Gathara · GitHub ↗ · v1.0.0
cross-platform ✓ Security Clean
53037
Downloads
329
Stars
561
Active Installs
1
Versions
Install in OpenClaw
/install 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.
Usage Guidance
Before installing, make sure you are comfortable with the skill running yt-dlp locally and making network requests for video URLs you ask it to process. Install yt-dlp only from a trusted package source, and treat fetched transcript content as untrusted third-party text rather than instructions for your agent.
Capability Analysis
Type: OpenClaw Skill Name: youtube-watcher Version: 1.0.0 The OpenClaw skill 'youtube-watcher' is designed to fetch and clean YouTube video transcripts using the `yt-dlp` utility. The `SKILL.md` provides clear instructions for the agent, and the `scripts/get_transcript.py` script executes `yt-dlp` with specific arguments to download subtitles into a temporary directory, then cleans and prints the text. There is no evidence of data exfiltration, malicious execution beyond the intended use of `yt-dlp`, persistence mechanisms, prompt injection attempts against the agent, or obfuscation. The behavior is fully aligned with its stated purpose.
Capability Assessment
Purpose & Capability
The stated purpose is fetching YouTube transcripts for summarization, Q&A, and extraction; the script calls yt-dlp to download subtitles, cleans VTT text, and prints the transcript.
Instruction Scope
The runtime examples are user-directed and transcript-focused. Some trigger phrases such as "summarize video" and "analyze video" are broad, but the artifact context consistently frames use around YouTube transcripts.
Install Mechanism
The skill requires the external yt-dlp binary and discloses brew and pip install options. Users should trust the package source they choose.
Credentials
The skill runs a local subprocess and contacts the video source for the user-provided URL, which is proportionate to transcript retrieval. The command is invoked as an argument list, not through shell string interpolation.
Persistence & Privilege
The script uses a temporary directory for subtitle files and contains no background service, credential access, privileged operation, broad local indexing, or persistent storage.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install youtube-watcher
  3. After installation, invoke the skill by name or use /youtube-watcher
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of youtube-watcher. - Fetch and read transcripts from YouTube videos. - Summarize videos, answer questions, or extract specific information from transcripts. - Trigger phrases include: "watch youtube", "summarize video", "video transcript", "youtube summary", and "analyze video". - Requires installation of `yt-dlp` for transcript retrieval. - Only works with videos that have closed captions or auto-generated subtitles.
Metadata
Slug youtube-watcher
Version 1.0.0
License
All-time Installs 1645
Active Installs 561
Total Versions 1
Frequently Asked Questions

What is 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. It is an AI Agent Skill for Claude Code / OpenClaw, with 53037 downloads so far.

How do I install YouTube Watcher?

Run "/install youtube-watcher" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is YouTube Watcher free?

Yes, YouTube Watcher is completely free (open-source). You can download, install and use it at no cost.

Which platforms does YouTube Watcher support?

YouTube Watcher is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created YouTube Watcher?

It is built and maintained by Michael Gathara (@michaelgathara); the current version is v1.0.0.

💬 Comments