Instagram Downloader

Instagram Media Downloader

Downloads photos and videos from public Instagram posts, reels, and carousels using gallery-dl. Carousel posts are split into one file per image/video automatically — no special-casing needed.

Requirements

• Outbound internet access to instagram.com. This will NOT work in network-restricted sandboxes (e.g. some cloud Cowork/Claude Code environments block instagram.com by org policy) — if a request hangs or returns a connection/proxy error, tell the user this environment can't reach Instagram rather than retrying repeatedly.
• gallery-dl (Python). Check first, install if missing:

  gallery-dl --version || pip install --upgrade gallery-dl --break-system-packages
  

  (drop --break-system-packages on systems where pip doesn't require/support it)

Steps

1. Confirm the URL is a public Instagram post/reel/carousel, e.g. https://www.instagram.com/p/XXXXXXXXX/ or https://www.instagram.com/reel/XXXXXXXXX/. Private accounts/posts are out of scope for this skill (no login support).

2. Pick an output directory. Default to ./instagram-downloads/<shortcode> (the shortcode is the id segment after /p/ or /reel/ in the URL) unless the user specifies a destination.

3. Run gallery-dl:

   gallery-dl --directory "<output_dir>" "<instagram_url>"
   

   • Single photo/video post → downloads one file.
   • Carousel → gallery-dl walks every sidecar item and writes each as a separate numbered file (e.g. ..._1.jpg, ..._2.jpg, ..._3.mp4) in the same folder automatically.
   • Reels and video posts save as .mp4.

4. If the command fails:

   • 401 / login wall → the post is private or Instagram is demanding auth; this skill only handles public content — tell the user and stop.
   • 429 / rate limited → wait and retry, or add --sleep 5-15 between requests for multiple URLs.
   • "Unsupported URL" / empty output with no clear error → Instagram changes its markup often and breaks scrapers; try pip install -U gallery-dl --break-system-packages first, since this is almost always an outdated extractor, not a bad URL.

5. Verify and report: list the output directory (ls -la <output_dir>) and tell the user what was downloaded (file count, e.g. "3 images + 1 video from the carousel"). In Cowork, deliver the files with SendUserFile; on a local/CLI environment, just confirm the saved path.

Notes

• Only download content the user has the right to reuse (their own posts, or with explicit permission) — respect Instagram's Terms of Service and the original poster's rights, especially before reusing downloaded media elsewhere (e.g. in GuideMapper content).
• Out of scope: Stories and Highlights (expire, need login) and private accounts. If the user needs those, that's a separate, login-based extension of this skill.