菜单
Intelligent video processor for downloading media and extracting transcripts from YouTube and 1000+ supported sites. Automatically handles format selection, subtitle extraction, and post-processing.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | video-toolkit |
| description | Intelligent video processor for downloading media and extracting transcripts from YouTube and 1000+ supported sites. Automatically handles format selection, subtitle extraction, and post-processing. |
| allowed-tools | Read Write Glob Grep Bash(yt-dlp:*) Bash(ffmpeg:*) Bash(which:*) Bash(python3:*) |
This skill provides intelligent capabilities to download video/audio and extract transcripts using yt-dlp.
Goal-Oriented Workflow: DO NOT ask for confirmation unless the user's intent is ambiguous.
en).--cookies-from-browser chrome) ONLY if the download fails with "Sign in" or "403 Forbidden" errors.mp4 for video, srt -> txt for transcripts.Best for extracting content for reading, summarizing, or RAG.
Extract Subtitles: Downloads subtitles (manual or auto-generated) without downloading the video.
yt-dlp --write-sub --write-auto-sub --sub-lang "en,zh-Hans,.*" --convert-subs srt --skip-download -o "%(title)s.%(ext)s" "[URL]"
(Note: Adjust --sub-lang priority based on the video's likely language if known)
Convert to Text:
Run the cleaning script to produce a readable .txt file:
python3 scripts/clean_transcript.py "[FILE].srt"
Best for saving files locally.
Standard Video:
yt-dlp -o "%(title)s.%(ext)s" "[URL]"
Audio Only:
yt-dlp -x --audio-format mp3 -o "%(title)s.%(ext)s" "[URL]"
yt-dlp --cookies-from-browser chrome ... (or firefox/safari/edge depending on user OS).