agentleFS
Sign inStart free
titouannwtttitouannwtt/claude-tools/skills/media-download/SKILL.md

media-download

Skill1 starsChanged 12 days agoMITView on GitHub

Download the video, or only the audio as MP3, from YouTube (videos, Shorts), Instagram (posts, Reels), TikTok and ina.fr links, or read their metadata and the YouTube thumbnail. Use when the user gives a YouTube, Shorts, Instagram, Reel, TikTok or INA link, or a list of links, and asks to download, extract the audio, keep an MP3 or build an audio bank. For the transcript, use video-transcript.

  • Reads credentials
  • Installs packages

4 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Update yt-dlp (pip install -U yt-dlp) and rerun.

    line 50

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

media-download

A script relies on yt-dlp (download) and ffmpeg (MP3 conversion, mp4 merge). It handles a list of links in one call. Reserve it for content the user has the right to download (their own, free license, authorized use): if the doubt is serious, say so rather than executing.

Commands

Script: ~/.claude/skills/media-download/scripts/media_download.py. The last stdout line is JSON.

  • doctor: checks yt-dlp, ffmpeg and the YouTube, Instagram, TikTok and INA extractors. Run it first on a new machine.
  • audio <url> [<url> ...] [--liste FILE] --dossier DIR [--cookies FILE] [--qualite 192]: for each link, the best audio stream converted to MP3 (192 kb/s by default). Writes DIR/<slug>.mp3 and DIR/<slug>.json; every video or intermediate file is deleted. --liste is a text file with one link per line (# comments allowed).
  • video <url> [<url> ...] [--liste FILE] --dossier DIR [--height 1080] [--max-filesize 500M]: mp4 video, same outputs.
  • info <url>: title, author, date, duration, platform, available heights, license. Run before a long download.
  • thumbnail <url> --dossier DIR: largest YouTube thumbnail (maxres, then sd, hq, mq).

audio and video end with {"fichiers": [paths], "echecs": [{"lien", "erreur"}], "metadonnees": [...]}. One failing link never stops the others. <slug>.json holds fichier, titre, auteur, date (YYYY-MM-DD or null), plateforme (YouTube, Instagram, TikTok, INA), lien (canonical), duree_s, id. The slug is a short ASCII title (the id is added on collision). Tell the user where the files are, and report each entry of echecs.

Platforms and login

  • YouTube (watch, Shorts, youtu.be) and INA: no account needed.
  • Instagram needs a session, TikTok sometimes: pass --cookies FILE (Netscape format, exported from a browser by the user; a secondary account is advised). A "login required" failure says so in its message. The script never reads the browser cookies by itself: do not add --cookies-from-browser unless the user asks for it.
  • auteur can be null (INA gives no channel).

Exit codes (audio, video)

CodeMeaningAction
0At least one link succeededGive the paths, list the failures.
2No valid linkAsk for correct links.
5Missing dependency (yt-dlp, ffmpeg) or proxy not configuredInstall it (pip install -U yt-dlp, brew install ffmpeg), run doctor.
1Every link failedRead echecs: login (use --cookies), blocked (below), or unexpected (repair).

info and thumbnail use the same codes plus 4 (blocked) and 8 (no thumbnail: private or deleted video).

Blocked (429, 403, "not a bot")

  1. Update yt-dlp (pip install -U yt-dlp) and rerun. 2. Give --cookies FILE. 3. Only once, --use-proxy (residential proxy from local.env, key IPROYAL_RESIDENTIAL_URL, limited budget of 2 GB): prefer audio with --max-filesize, avoid video. Never display the proxy value.

Repair

Fix <claude-tools>/skills/media-download/scripts/media_download.py, then run python3 -m unittest discover -s tests in the skill's folder. Two attempts at most, then report the exact message.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free