youtube-scribelisted

# YouTube Scribe Given a YouTube URL, produce a readable transcript and a TL;DR. This skill's directory is referred to as `{baseDir}`. It contains `scripts/yt-transcript`, a thin wrapper around the separate **baoyu-youtube-transcript** engine. ## Prerequisites The engine is a separate skill and is not bundled. If it is missing, the wrapper prints the install command and stops. Install it once with: ``` npx skills add JimLiu/baoyu-skills --skill baoyu-youtube-transcript -g ``` `bun` (preferred) or `npx` is required to run the engine. ## Steps 1. Extract the YouTube URL from the user's request. 2. Run the wrapper from this skill's directory, single-quoting the URL (`?` is a shell glob): ``` {baseDir}/scripts/yt-transcript '<youtube-url>' ``` If `{baseDir}` cannot be resolved but `yt-transcript` is on `PATH`, that works too. Useful flags: `--out-dir <dir>` (default `./transcripts`), `--languages en,zh` (priority order), `--refresh` (ignore engine cache). If the engine is missing, show the install command the wrapper prints and stop. 3. The command prints the transcript file path (last line of stdout). Read that file. 4. Write `summary.md` in the same directory as the transcript: - `# TL;DR` — 3-6 bullets capturing what the video actually says. - `## Key points` — the main arguments in order. - `## Notable details` — specific facts, numbers, and names worth knowing. Base it on the transcript content, **not** the video description.