Social media downloader

Skill TikHub/tikhub-plugin/skills/social-media-downloader

Download video, audio, or images (no-watermark where available) from a social-media URL or list of URLs. Use when the user pastes a TikTok/Douyin/Instagram/YouTube/Twitter/Xiaohongshu link and wants the media file, or says "download this video", "save without watermark", "grab the audio". Dispatches to the right per-platform TikHub endpoint.From its SKILL.md

Install

npx -y skills add TikHub/tikhub-plugin --skill social-media-downloader

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

5 stars5 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

3.5 KB, as published. Nobody here has run it

Social Media Downloader

Turn a post URL into a saved media file by routing it to the correct platform endpoint, extracting the media URL from the response, and downloading it.

Setup gate

[ -z "${TIKHUB_API_KEY:-}" ] && echo "Set TIKHUB_API_KEY first (see tikhub-onboarding)."

Step 1 — Detect the platform from the URL and pick the endpoint

Platform	Endpoint	Input	Media field
TikTok	`GET /api/v1/tiktok/app/v3/fetch_one_video_by_share_url`	`share_url`	no-watermark `play_addr` / `download_addr` `url_list`
TikTok (by id)	`GET /api/v1/tiktok/app/v3/fetch_one_video`	`aweme_id`	`url_list`
Douyin	`GET /api/v1/douyin/app/v3/fetch_one_video_by_share_url`	`share_url`	video `url_list`
Douyin (by id)	`GET /api/v1/douyin/app/v3/fetch_one_video_v2`	`aweme_id`	video `url_list`
YouTube	`GET /api/v1/youtube/web_v2/get_video_streams_v2`	`video_id` \| `video_url`	stream URLs (pick resolution)
Instagram	`GET /api/v1/instagram/v2/fetch_post_info`	`code_or_url`	media URL(s)
Xiaohongshu	`GET /api/v1/xiaohongshu/app_v2/get_video_note_detail` (video) / `get_image_note_detail` (images)	`note_id`	video / image URLs
Twitter/X	`GET /api/v1/twitter/web/fetch_tweet_detail`	`tweet_id`	media `url_list`

For TikTok/Douyin, the *_by_share_url endpoints take the raw URL directly — no need to extract an id. For the others, pull the id/code from the URL (or resolve it via the platform skill).

Step 2 — Call the endpoint

# TikTok by share URL (simplest — pass the URL straight through)
URL="https://www.tiktok.com/@nasa/video/7650608519288245534"
curl -s "https://api.tikhub.io/api/v1/tiktok/app/v3/fetch_one_video_by_share_url?share_url=$(python3 -c "import urllib.parse,sys;print(urllib.parse.quote(sys.argv[1],safe=''))" "$URL")" \
  -H "Authorization: Bearer $TIKHUB_API_KEY" -o /tmp/media.json

# YouTube streams
curl -s "https://api.tikhub.io/api/v1/youtube/web_v2/get_video_streams_v2?video_id=dQw4w9WgXcQ" \
  -H "Authorization: Bearer $TIKHUB_API_KEY" -o /tmp/media.json

Step 3 — Extract the media URL, then download

Parse the response for the highest-quality (no-watermark, for TikTok/Douyin) media URL — usually a url_list array or a stream URL — then save it:

curl -L "<media_url_from_response>" -o video.mp4
file video.mp4   # confirm it's a real media container

Batch downloads

Loop over a URL list, one at a time with a small delay (respect QPS 10/sec). Each parse is one billed call — warn the user for large batches and hand off to bulk-data-export for big jobs.

Verification gate

Response is code: 200 and contains a non-empty media URL.
Downloaded file is non-empty and the expected type (file video.mp4 shows a video container).
Not an auth/credit error.

Red flags

Claiming a download succeeded without checking the file is non-empty/valid.
Using the wrong platform endpoint for a URL — detect the platform first.
Unbounded batch loops (credit burn + rate limits).
Re-hosting/redistributing copyrighted media — respect platform ToS and the user's rights.