Twitter Downloader Skill

Download and summarize Twitter/X posts, save media into Minis, and return chat-ready Markdown.

When to Use

• User provides a twitter.com or x.com status URL.
• User asks to download Twitter/X images, GIFs, or videos.
• User asks what a tweet says/contains.
• User wants downloaded media inserted/displayed in chat as Markdown.

What It Does

1. Parses username and tweet/status ID from Twitter/X URL variants.
2. Fetches structured JSON from api.fxtwitter.com, with fallback to api.vxtwitter.com.
3. Generates a short summary:
   • author
   • text
   • created time if available
   • sensitive flag
   • original media URLs
4. Downloads images, video thumbnails, and GIF/video files by default.
5. Returns Markdown containing:
   • ## Tweet Summary
   • summary/raw JSON links
   • inline media syntax for images/thumbnails/videos/GIFs: ![filename](minis://...)

Dependencies

• curl
• jq
• python3

The helper script auto-installs missing packages with apk add --no-cache.

Helper Script

Path: /var/minis/skills/twitter-downloader/scripts/twitter_downloader.sh

Usage:

/var/minis/skills/twitter-downloader/scripts/twitter_downloader.sh "<tweet_url>"

Options:

--dir DIR        Output directory, default /var/minis/workspace/tweet_media
--images         Download images/thumbnails only in addition to summary
--video          Download videos/GIFs only in addition to summary
--all            Download images/thumbnails and videos/GIFs; default when no media flag is provided
--no-download    Only fetch summary/JSON and return Markdown links for those files
--json-only      Fetch and print raw tweet JSON only; no Markdown output

Examples:

# Default: download all available media and output Markdown
/var/minis/skills/twitter-downloader/scripts/twitter_downloader.sh "https://x.com/user/status/123"

# Summary only, no media downloads
/var/minis/skills/twitter-downloader/scripts/twitter_downloader.sh "https://x.com/user/status/123" --no-download

# Custom output directory
/var/minis/skills/twitter-downloader/scripts/twitter_downloader.sh "https://x.com/user/status/123" --dir "/var/minis/workspace/tweet_media"

Generated files:

/var/minis/workspace/tweet_media/<tweet_id>.json
/var/minis/workspace/tweet_media/<tweet_id>_summary.txt
/var/minis/workspace/tweet_media/<tweet_id>/<media files>

Agent Workflow

1. Run the helper script with the tweet URL.
2. Paste stdout directly into chat.
3. Do not merely mention the folder path; include generated Markdown links.
4. Keep images, thumbnails, videos, and GIFs as inline media syntax: ![filename](minis://...)
5. Keep JSON/text summary files as normal links: [filename](minis://...)
6. If Sensitive: True, preserve that field and avoid adding explicit extra descriptions unless the user asks.

Output Format

The helper outputs Markdown similar to:

## Tweet Summary

- Author: ...
- Text: ...
- Created: ...
- Sensitive: false
- Media:
-   photo https://...
- Downloaded images: 1, videos: 1

## File Links

- [summary.txt](minis://workspace/...)
- [raw.json](minis://workspace/...)

## Media

![photo.jpg](minis://workspace/.../photo.jpg)

![video.mp4](minis://workspace/.../video.mp4)

Error Handling

• If URL parsing fails, ask for a valid twitter.com/x.com status URL.
• If both APIs fail, report that the tweet may be private/deleted or the API may be temporarily unavailable.
• If no media is found, still return the tweet summary and JSON/summary links.

Notes

• Video/GIF downloads choose the best bitrate variant when available.
• Video/GIF thumbnails are downloaded when image download is enabled.
• Minis URLs are percent-encoded by the helper.