YouTube Transcript Generator

Download and generate clean, readable transcripts from any YouTube video. Extracts subtitles (auto-generated or manual), removes timestamps and formatting, a...

Security Scan

The skill appears to do exactly what it claims — download and clean YouTube subtitles using yt-dlp and a bundled shell script — with no unexpected credential requests or hidden network endpoints.

Purpose & Capability

The skill's name/description match its behavior: the SKILL.md and the bundled script download subtitles and produce cleaned transcripts. Minor inconsistency: the registry metadata declares no required binaries, but the SKILL.md and script explicitly require yt-dlp (and use python3).

Instruction Scope

Runtime instructions and the script stay within the stated purpose: calling yt-dlp to fetch subtitle files, cleaning them, and writing a transcript file. The script only reads the downloaded subtitle file(s) and writes output to the current directory / a temp dir; it does not attempt to read unrelated system files or environment variables.

Install Mechanism

This is an instruction-only skill with a local shell script; there is no install spec and nothing is downloaded by the skill itself. The only external dependency is yt-dlp (recommended install via brew or pip) which the user must install separately.

Credentials

The skill requests no environment variables or credentials and does not require access to other services beyond YouTube via yt-dlp. The lack of declared required binaries in registry metadata is the only proportionality mismatch (yt-dlp/python3 are required in practice).

Persistence & Privilege

The skill is not always-enabled and does not modify other skills or system-wide configs. It runs on demand and writes only its own output file; no elevated privileges are requested.

Assessment

This skill is coherent and appears to do what it says, but check a few practical items before use: (1) Install yt-dlp from an official source (brew, pip, or the project's releases) because the script depends on it — the registry metadata omitted this requirement. (2) Review the bundled scripts (you already have them) and run them in a safe working directory; the script writes transcript_<VIDEO_ID>.txt to your current directory and uses a temporary folder for subtitles. (3) Run as a non-root user and try it on a harmless public video first to confirm behavior. (4) Be aware that yt-dlp will contact YouTube (network activity) and that downloading subtitles may have copyright/legal considerations in your jurisdiction. (5) Verify any external links in the SKILL.md before following them. If you want higher assurance, ask the publisher for provenance (source/homepage) or request the skill be updated so metadata lists yt-dlp/python3 as required binaries.

Download clean transcripts from any YouTube video URL.

Requirements

• yt-dlp must be installed (brew install yt-dlp or pip install yt-dlp)

Usage

Run the bundled script with a YouTube URL:

bash scripts/get_transcript.sh "https://www.youtube.com/watch?v=VIDEO_ID"

The script will:

1. Try to download English manual subtitles first
2. Fall back to auto-generated English subtitles
3. Try all available languages if English is unavailable
4. Clean the raw subtitle file into readable paragraphs
5. Output the transcript to stdout and save to transcript_VIDEO_ID.txt

Options

# Save to a specific file
bash scripts/get_transcript.sh "URL" output.txt

# Get transcript WITH timestamps (default: without)
bash scripts/get_transcript.sh "URL" output.txt en timestamps

# Get transcript in a specific language
bash scripts/get_transcript.sh "URL" output.txt fr

How It Works

1. yt-dlp downloads the subtitle track (VTT/SRT format)
2. The script strips HTML tags and duplicate lines
3. Without timestamps (default): merges into clean, readable paragraphs
4. With timestamps: preserves [HH:MM:SS] markers before each line for easy reference

Example Output

Input: https://www.youtube.com/watch?v=HMTxOecbyPg

Output:

How OpenClaw Runs My Entire Business. I record a podcast episode and that is
literally the only thing I do. Everything else is handled by 13 AI agents
running on a Mac Mini in my office...

Troubleshooting

• No subtitles found: Not all videos have subtitles. The script will report which languages are available.
• yt-dlp not found: Install with brew install yt-dlp (macOS) or pip install yt-dlp.
• Rate limited: Wait a moment and retry. YouTube occasionally throttles subtitle requests.

Links

• Full guides and templates: OpenClaw Lab
• Free OpenClaw installer: installopenclawnow.com
• Community: OpenClaw Lab on Skool