YouTube video summarizer with speaker detection, formatted documents, and audio output. Works out of the box with macOS built-in TTS. Optional recommended tools (pandoc, ffmpeg, mlx-audio) enhance quality. Requires internet for YouTube access. No paid APIs or subscriptions. Use when user sends a YouTube URL or asks to summarize/transcribe a YouTube video.
Security Analysis
high confidenceTubeScribe's declared capabilities, required files, and runtime instructions are consistent with a local YouTube summarizer; nothing in the package requests unrelated credentials or silent external endpoints.
Name/description (YouTube summarizer + TTS) matches the code and declared requirement. The only required binary declared in metadata is 'summarize', which the code and README reference; optional tools (pandoc, ffmpeg, yt-dlp, mlx-audio, kokoro) are documented as optional and used for better output. No unexpected credentials or unrelated binaries are required.
SKILL.md instructs the agent to spawn a sub-agent that runs the local pipeline (scripts/tubescribe.py) and to read/write local files (Source/Output paths, ~/.tubescribe/config.json, ~/Documents/TubeScribe). That is appropriate for this skill. Note: SKILL.md also emphatically forbids installs/network calls beyond YouTube for spawned sub-agents, but the included setup.py (meant for interactive first-time setup) will perform downloads from well-known hosts if you run it manually — the policy difference between automated sub-agents and manual setup is intentional but worth noting.
There is no automated install spec (instruction-only in registry) — the skill includes local Python scripts. setup.py can download official releases (pandoc from GitHub, etc.) and pip packages for optional tooling; those downloads are from well-known sources (GitHub, PyPI). This is moderate-risk only when you choose to run setup.py; the runtime pipeline itself doesn't require automatic downloads and the SKILL.md instructs sub-agents not to install software.
The skill asks for no environment variables or credentials. It reads and writes user-local config (~/.tubescribe/config.json), temp files, and outputs to ~/Documents/TubeScribe — all appropriate for a tool that generates documents and audio. No unrelated secrets or system-wide credentials are requested.
The skill is not always-enabled and does not request elevated privileges. It instructs spawning a sub-agent to run the pipeline (normal for async processing). It does not modify other skills or system-wide agent settings. Running setup.py may create ~/.openclaw/tools/ entries and venvs for optional TTS tooling, but that's local and user-initiated.
Guidance
TubeScribe appears to be what it says: a local YouTube summarizer that reads/writes files in your home directory and uses YouTube/network access only to fetch video data. Before installing/using: 1) Verify you have or want the required 'summarize' CLI (the skill will not auto-install it). 2) If you run setup.py interactively, expect downloads from GitHub/PyPI/Homebrew for optional tools (pandoc, yt-dlp, mlx-audio, Kokoro); only run setup if you trust those sources. 3) The skill will create/read ~/.tubescribe/config.json and output to ~/Documents/TubeScribe — if you prefer isolation, run it in a dedicated account/container. 4) If you are concerned about supply-chain risk, inspect scripts/setup.py and the included Python modules before running setup or installing optional dependencies.
Latest Release
v1.1.8
- Description updated: works out of the box, optional tools enhance quality, internet required for YouTube - Added OpenClaw metadata: declares summarize as required binary - Added security clarifications: no data uploaded, sub-agent has strict no-install instructions - Expanded setup.py documentation in README - Fixed display name on ClawHub: TubeScribe (was Tubescribe)
More by @matusvojtek
Published by @matusvojtek on ClawHub
