← Back to Skills Marketplace
383
Downloads
0
Stars
1
Active Installs
1
Versions
Install in OpenClaw
/install youtube-to-chinese
Description
将提供的 YouTube 视频链接转录成完整中文稿，包含内容摘要和视频核心亮点，便于快速理解和复习。
Usage Guidance
This skill appears coherent and implements a reasonable workflow for converting YouTube videos to Chinese transcripts. Before using it, consider: 1) Only supply cookies (browser:chrome or cookies.txt) if you understand that this gives yt-dlp access to your browser session cookies — avoid on sensitive accounts. 2) Verify any /tmp/yt-dlp binary before running (the script will prefer /tmp/yt-dlp if present). 3) If you run the suggested curl to update yt-dlp, ensure you trust the GitHub URL (the script points to the official yt-dlp release). 4) audios_understand is claimed to be an internal tool that processes local files; confirm your platform's privacy policy for how transcriptions are handled. If you avoid passing browser cookies and only run this on non-sensitive accounts/machines, the risk is low.
Capability Analysis
Type: OpenClaw Skill
Name: youtube-to-chinese
Version: 1.0.0
The skill bundle contains high-risk capabilities, specifically the instruction to download and execute an external binary (yt-dlp) from GitHub and the use of the `--cookies-from-browser` flag in `scripts/download_audio.sh`, which can access sensitive session data. While these are functional requirements for bypassing YouTube's bot detection, they provide a path for RCE and credential theft. Additionally, the execution of shell scripts with user-provided URLs in `SKILL.md` poses a command injection risk to the agent's host environment.
Capability Assessment
Purpose & Capability
Name/description (YouTube → Chinese transcript/summary) aligns with included files and runtime steps: a downloader script and instructions to pass the downloaded audio to the platform's audios_understand tool. No unrelated binaries, env vars, or hosts are requested.
Instruction Scope
SKILL.md stays within scope: download audio, transcribe via audios_understand, and format output. It does recommend optional use of browser cookies (browser:chrome) or a cookies.txt export to bypass YouTube bot/region checks — this increases the scope to reading browser cookies if the user explicitly chooses that option.
Install Mechanism
Instruction-only skill with a small shell script. No install spec. The only external download suggested (if needed) is yt-dlp from the official GitHub releases URL — a well-known release host. No obscure URLs or archive extraction are used.
Credentials
The skill itself requires no environment variables or credentials. The only potential sensitive access is optional: using browser:chrome or passing a cookies.txt file to allow yt-dlp to use logged-in cookies. That is user-controlled but can expose browser session cookies if used.
Persistence & Privilege
always is false and the skill does not request persistent system privileges or modify other skills. It runs a local script and relies on local binaries (yt-dlp, ffmpeg if user chooses to split).
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install youtube-to-chinese - After installation, invoke the skill by name or use
/youtube-to-chinese - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release: YouTube视频转中文稿，含摘要和亮点
Metadata
Frequently Asked Questions
What is YouTube to Chinese?
将提供的 YouTube 视频链接转录成完整中文稿，包含内容摘要和视频核心亮点，便于快速理解和复习。 It is an AI Agent Skill for Claude Code / OpenClaw, with 383 downloads so far.
How do I install YouTube to Chinese?
Run "/install youtube-to-chinese" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is YouTube to Chinese free?
Yes, YouTube to Chinese is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does YouTube to Chinese support?
YouTube to Chinese is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created YouTube to Chinese?
It is built and maintained by 古 晖 (@webguhui); the current version is v1.0.0.
More Skills