Lyt
Skill TanaTTV/lyt/skills/lyt
Download permitted audio or video from YouTube and other yt-dlp-supported sites with the lyt CLI. Use for media downloads, MP3 extraction, quality selection, clips, format inspection, and exact local output paths.From its SKILL.md
npx -y skills add TanaTTV/lyt --skill lytAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
2.4 KB, as published. Nobody here has run it
Use lyt for local media
lyt is a local CLI around yt-dlp and ffmpeg. Only download media the user owns
or has permission to use. A public URL does not automatically grant permission.
Find or install the command
- Try
lyt --version. - In the lyt repository, use
node bin/lyt.js. - Otherwise ask before globally installing
@tanattv/lyt.
Global npm installation and first-use managed binary downloads are side effects. Do not perform either without explicit user approval. Check capabilities with:
lyt doctor --json
Inspect before downloading. These commands have no download side effects:
lyt info --json "URL" # media metadata and formats (schema lyt.info.v1)
lyt capabilities --json # commands, flags, and schemas (lyt.capabilities.v1)
lyt info describes what a URL offers (title, duration, formats); lyt doctor
describes what the local environment can do. Use lyt capabilities to discover
the CLI surface instead of parsing human help text.
Always quote URLs. Prefer --dry-run --json before a large or uncertain job.
For bounded agent calls, --json emits one JSON document using schema lyt.result.v1 on stdout; setup and progress diagnostics go to stderr.
lyt --audio --dry-run --json "URL"
lyt --audio --json "URL"
lyt --mp3 -q 192K --json "URL"
lyt --video -q 1080p --max-filesize 2G --json "URL"
lyt --clip 1:10-2:45 --mp3 --json "URL"
lyt --list-formats --json "URL"
Files go to ./downloads under the current working directory unless -o is
supplied. Before a real job, confirm the intended output directory when it is
material to the user's workflow.
Keep these behaviors opt-in:
--playlist--force-overwrite--redownload- browser cookies or authentication material
- external downloaders
- managed tool installation
Never use cookies, tokens, private URLs, or authentication material unless the user explicitly requests it. Never include those values in logs or bug reports.
After success, read results[].files and report those exact paths. If a result
is skipped, explain whether a matching artifact already exists or a size guard
was triggered. Use --redownload only when the user asks for another copy.