Case 02821
A fast, offline static risk analysis CLI for AI agent skill files. Detects malicious instructions, steganographic payloads, and dangerous capability chains.
npx -y skills add knownasnaffy/prompthound --skill case_02821Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Download videos and get transcripts, summaries, or metadata from YouTube, TikTok, Instagram, and X (Twitter). Use when the user shares a video URL and wants a transcript, summary, key points, quotes, or to download the video. Triggers on phrases like "transcript this video", "summarize this YouTube video", "what does this video say", "download this TikTok", "get captions from", or any video URL from youtube.com, youtu.be, tiktok.com, instagram.com, x.com, twitter.com. Requires: yt-dlp, python3, curl. Optional: ffmpeg (for TikTok/Instagram/X audio), OPENAI_API_KEY (Whisper fallback — uploads audio to OpenAI).
SKILL.md
3.2 KB, 665 tokens by cl100k_base, as published. Nobody here has run it
video-intel
Downloads videos and extracts transcripts using yt-dlp (captions) with OpenAI Whisper fallback.
Required Dependencies
| Dependency | Purpose | Required? |
|---|---|---|
yt-dlp | Fetch captions and download audio/video | ✅ Always |
python3 | Parse VTT/SRT caption files | ✅ Always |
curl | Call OpenAI Whisper API | ✅ For Whisper fallback |
ffmpeg | Extract audio from TikTok/Instagram/X | ⚠️ Non-YouTube only |
OPENAI_API_KEY | Authenticate with OpenAI Whisper API | ⚠️ Only if captions unavailable |
Install binaries:
# yt-dlp
curl -L https://github.com/yt-dlp/yt-dlp/releases/latest/download/yt-dlp -o ~/bin/yt-dlp && chmod +x ~/bin/yt-dlp
# ffmpeg (Debian/Ubuntu)
sudo apt install ffmpeg
⚠️ Privacy Notice — OpenAI Audio Upload
When captions are unavailable (common for TikTok, Instagram, X), the script downloads the audio and uploads it to OpenAI's transcription API (https://api.openai.com/v1/audio/transcriptions). This means:
- Audio content leaves your machine and is sent to OpenAI
- Requires
OPENAI_API_KEYto be set - If you don't want external transmission: don't set
OPENAI_API_KEY, or use a local transcription model
YouTube videos almost always have captions and will not trigger an upload.
Script
~/.openclaw/skills/video-intel/scripts/video-intel.sh
Workflows
Get transcript
~/.openclaw/skills/video-intel/scripts/video-intel.sh transcript <url>
- YouTube: uses built-in captions/auto-subs (fast, no audio download or external upload)
- TikTok/Instagram/X: downloads audio → uploads to OpenAI Whisper for transcription
- Preferred language:
--lang ptfor Portuguese
Get video info
~/.openclaw/skills/video-intel/scripts/video-intel.sh info <url>
List available caption tracks
~/.openclaw/skills/video-intel/scripts/video-intel.sh captions <url>
Download video
~/.openclaw/skills/video-intel/scripts/video-intel.sh download <url> [--format audio|720p|best]
After getting transcript
- Summary: Summarize in 3-5 bullet points
- Key quotes: Extract most notable quotes
- Full summary: Write a paragraph summary with context
- Translation: Translate to the user's language if different
Notes
- YouTube auto-captions are usually available even without ffmpeg
- TikTok/Instagram/X require ffmpeg for audio extraction
- Large videos (>25MB audio) may hit OpenAI's file size limit — use
--format audioto get a smaller mp3 - Output cached in
/tmp/video-intel/by default