メニュー
Downloads YouTube subtitles and summarizes the video content using Gemini models while filtering out sponsor promotions. Use when the user asks to summarize, analyze, or get details from a YouTube URL.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Customizes the Antigravity CLI status bar to show folder icons, home-abbreviated current working directory, model name, and 5H/7D token quota usage with live refresh countdown.
Use this skill to parse complex documents (PDFs, Word documents, PowerPoint presentations, Excel spreadsheets, or images) into clean Markdown or structured JSON using the LlamaParse API. Make sure to use this skill whenever the user asks to extract tables from PDFs, handle complex document structures (multi-column, charts, forms), parse files with custom prompt guidelines, or convert documents to LLM-ready markdown, even if they don't explicitly mention 'LlamaParse'.
Use this skill to convert docx and other office documents to Markdown using markitdown, or perform document content analysis. Triggers include: markitdown, convert docx to md, convert word to markdown, document extraction.
Enforce, validate, lint, and manage Conventional Commits across git repositories. Generates semantic changelogs and installs git hooks. Use this skill whenever the user mentions commit messages, commit conventions, conventional commits, semantic versioning, changelogs, commit linting, git hooks for commits, or wants to standardize their commit format — even if they don't explicitly say "conventional commits".
Generates images and text via reverse-engineered Gemini Web API. Suitable for AI Pro and higher tier subscription users. Supports text generation, image generation from prompts, reference images for vision input, and multi-turn conversations. Use when other skills need image generation backend, or when user requests "generate image with Gemini", "Gemini text generation", or needs vision-capable AI generation.
Posts content, deletes tweets, or exports/saves tweets to YAML on X (Twitter) using Chrome Computer Use Mode. Use when user requests "post to Twitter", "tweet", "share on X", "post to X", "delete tweet", "export tweets", or "save tweets".
| name | yt-summarize |
| description | Downloads YouTube subtitles and summarizes the video content using Gemini models while filtering out sponsor promotions. Use when the user asks to summarize, analyze, or get details from a YouTube URL. |
This skill allows downloading subtitles of YouTube videos and summarizing their contents using Gemini models. It filters out sponsor promotions and can start an interactive chat session about the video.
yt-dlp. No large video downloads are required.Ensure the following tools are installed on your system:
yt-dlp (e.g. brew install yt-dlp)python3httpx (pip3 install httpx)myopenai (a custom package/module present on your system)scripts/yt-summarize.py: The Python script that runs yt-dlp and calls the Gemini/LLM API.scripts/install.sh: Helper script to copy yt-summarize.py to the app data directory, grant execution permissions, and set up a command alias.<skill_dir>/scripts/install.sh
yt-summarize <youtube-url> [--lang <lang-code>] [--model <model-name>]
For example, to summarize a video:
yt-summarize https://www.youtube.com/watch?v=dQw4w9WgXcQ
To summarize with Chinese subtitles (if available):
yt-summarize https://www.youtube.com/watch?v=dQw4w9WgXcQ --lang zh-Hans
If you encounter zsh: no matches found: <url> error, it is because of the ? character in the YouTube URL. Wrap the URL in double quotes:
yt-summarize "https://www.youtube.com/watch?v=Rtkac4WHC1o"
If you see Sign in to confirm you’re not a bot, pass browser cookies using the --ytdlp-arg flag. On macOS, Chrome cookies are recommended:
yt-summarize "https://www.youtube.com/watch?v=Rtkac4WHC1o" --ytdlp-arg "--cookies-from-browser chrome"
(Note: Using --cookies-from-browser safari may fail due to macOS container sandbox permission issues.)
If the API key is not detected (e.g. for Gemini), you can specify another provider (like deepseek or openai) and pass the corresponding key:
yt-summarize "https://www.youtube.com/watch?v=Rtkac4WHC1o" --api deepseek