YouTube AnyCaption Summarizer
v1.1.3Turn YouTube videos into dependable markdown transcripts and polished summaries — even when caption coverage is messy. This skill works with manual closed ca...
⭐ 1· 78·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The name/description (YouTube summarizer that falls back to local Whisper) matches the required binaries (yt-dlp, ffmpeg, whisper-cli, python3), the Brew install lines, and the included scripts which fetch metadata, download subtitles/media, transcribe, normalize, and finalize summaries. Required items are proportional to the stated function.
Instruction Scope
SKILL.md instructs creation of a model directory (~/.openclaw/workspace), optionally downloading a Whisper model from Hugging Face, running yt-dlp (including optional --cookies or --cookies-from-browser) and local whisper-cli. These are expected for the stated workflow. Note: using --cookies-from-browser or providing a cookies file causes yt-dlp to access browser cookie data (sensitive) — the skill documents this and says to use only user-provided cookies. Also the scripts assume local filesystem read/write of per-video folders and generation/updates of markdown files (including workflow metadata JSON), which is normal for this tool.
Install Mechanism
Install instructions use Homebrew formulas for yt-dlp, ffmpeg, and whisper-cpp and an explicit curl from a Hugging Face model URL if missing. Homebrew formulas and the Hugging Face model URL are reasonable and expected here. The model download is external and large; the SKILL.md only downloads the model if absent, which is documented.
Credentials
The skill requests no environment variables or third-party credentials. The only sensitive capability is optional cookie support: the workflow supports --cookies and --cookies-from-browser to access restricted/private videos. That is justified by the use case but requires user caution because cookies can contain sensitive authentication tokens. No other unrelated credentials or config paths are requested.
Persistence & Privilege
The skill does not request always:true and is user-invocable only. It writes its own model directory (~/.openclaw/workspace) and per-video output folders, but it documents not touching OpenClaw config files. Nothing in the package attempts to modify other skills or global agent settings.
Assessment
This package appears to do what it says: transcribe and summarize YouTube videos using subtitle-first logic and local Whisper fallback. Before installing, note: 1) The workflow will download a Whisper model (large file) from Hugging Face into ~/.openclaw/workspace by default — ensure you have disk space and verify the URL if you need to be strict about sources. 2) If you need to process restricted/private videos, the tool accepts either a cookies file or --cookies-from-browser which can access browser cookie stores — only provide cookies you trust and be aware this can expose authentication tokens. 3) The scripts call external binaries (yt-dlp, ffmpeg, whisper-cli); install them from trusted package sources (brew) as suggested. 4) The package writes per-video folders and modifies markdown files in those folders; run it in a controlled parent directory if you want to limit where files are created. Overall, the requirements and behavior are proportional to the stated purpose.Like a lobster shell, security has layers — review code before you run it.
auto-captionsvk975pxe1wj8acrj4wbjvf9y2r183mtpbbatchvk975pxe1wj8acrj4wbjvf9y2r183mtpbclosed-captionsvk975pxe1wj8acrj4wbjvf9y2r183mtpblatestvk975pxe1wj8acrj4wbjvf9y2r183mtpbmarkdownvk979wb9mq32aay3cv1cakazmh583mxsjprivate-videosvk975pxe1wj8acrj4wbjvf9y2r183mtpbresearchvk979wb9mq32aay3cv1cakazmh583mxsjsubtitlesvk979wb9mq32aay3cv1cakazmh583mxsjsummarizationvk979wb9mq32aay3cv1cakazmh583mxsjtranscriptvk979wb9mq32aay3cv1cakazmh583mxsjvideo-analysisvk979wb9mq32aay3cv1cakazmh583mxsjwhispervk975pxe1wj8acrj4wbjvf9y2r183mtpbyoutubevk975pxe1wj8acrj4wbjvf9y2r183mtpbyoutube-summarizervk975pxe1wj8acrj4wbjvf9y2r183mtpbyoutube-transcriptvk975pxe1wj8acrj4wbjvf9y2r183mtpb
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
Binsyt-dlp, ffmpeg, whisper-cli, python3
Install
Install yt-dlp (brew)
Bins: yt-dlp
brew install yt-dlpInstall ffmpeg (brew)
Bins: ffmpeg
brew install ffmpegInstall whisper.cpp CLI (brew)
Bins: whisper-cli
brew install whisper-cpp