Instagram Post Downloader Skill

Downloads Instagram posts at full resolution from Instagram's CDN — no screenshots, no compression. Handles single images, carousels (multi-slide posts), and Reel cover images. For carousels, produces individual slide files plus a single stitched PDF. Supports batch URLs in one run.

PREREQUISITE — Domain Allowlist

Before this skill can fetch any media, you must add Instagram's CDN domain to Claude Code's allowlist:

Settings → Capabilities → Domain allowlist → Add:

*.cdninstagram.com

Without this, all CDN fetch calls will be blocked. If you see a permission error when Claude attempts a fetch to cdninstagram.com, this is the fix.

Required Inputs

Claude will ask for these if not provided upfront:

Batch input example:

https://www.instagram.com/p/ABC123/
https://www.instagram.com/p/DEF456/
https://www.instagram.com/p/GHI789/

Output Structure

For each URL processed, Claude creates a folder named after the post caption (first 40 characters, sanitised — spaces become underscores, special characters stripped). If no caption is available, the folder is named after the post shortcode.

Single image post

instagram-downloads/
└── this_is_the_caption_first_40_chars/
    ├── image.jpg
    └── metadata.txt

Carousel post

instagram-downloads/
└── carousel_caption_first_40_chars/
    ├── slide_01.jpg
    ├── slide_02.jpg
    ├── slide_03.jpg
    ├── slide_04.jpg
    ├── carousel.pdf          ← all slides stitched in order
    └── metadata.txt

Batch run (3 URLs)

instagram-downloads/
├── first_post_caption_sanitised/
│   ├── image.jpg
│   └── metadata.txt
├── second_post_carousel_caption/
│   ├── slide_01.jpg
│   ├── slide_02.jpg
│   ├── carousel.pdf
│   └── metadata.txt
└── third_post_caption_here/
    ├── image.jpg
    └── metadata.txt

metadata.txt format

Post URL:       https://www.instagram.com/p/XXXX/
Shortcode:      XXXX
Type:           carousel | single_image | reel
Slide count:    4  (carousel only)
Caption:        [full caption text]
Username:       @username
Fetched at:     2026-05-27T14:32:00Z
CDN URLs:
  slide_01.jpg  https://scontent.cdninstagram.com/v/...
  slide_02.jpg  https://scontent.cdninstagram.com/v/...

Completion summary (printed to terminal)

Instagram Post Downloader — Batch Complete
==========================================
URLs processed:   3
Posts saved:      3
Total files:      11  (9 images + 2 PDFs)
Skipped:          0
Output dir:       /Users/you/project/instagram-downloads/

Results:
  ✓ this_is_the_caption_first_40_chars/     1 image
  ✓ carousel_caption_first_40_chars/        4 slides → carousel.pdf
  ✓ third_post_caption_here/                1 image

How Claude Should Execute This Skill

Step 1 — Collect and validate inputs

1. Accept the URL(s) from the user. If the user pastes a comma-separated list, split on commas. If they paste one per line, split on newlines.
2. Validate each URL matches instagram.com/p/, instagram.com/reel/, or instagram.com/tv/. Flag malformed URLs before proceeding.
3. Confirm the output directory. If none provided, use ./instagram-downloads/ and tell the user.
4. Ask about PDF stitching preference only if the user hasn't said either way. Default is yes.

Step 2 — For each URL: fetch the post page

Fetch the Instagram post page HTML:

GET https://www.instagram.com/p/{shortcode}/?__a=1&__d=dis

Instagram frequently changes its API surface. Use this fallback chain in order:

Attempt A — JSON endpoint:

https://www.instagram.com/p/{shortcode}/?__a=1&__d=dis

Parse the JSON response. Look for graphql.shortcode_media or data.shortcode_media.

Attempt B — Embed page (most reliable):

https://www.instagram.com/p/{shortcode}/embed/captioned/

Fetch this page's HTML and extract og:image meta tags and any window.__additionalDataLoaded or window.__StaticData JSON blobs embedded in <script> tags.

Attempt C — oEmbed endpoint:

https://api.instagram.com/oembed/?url=https://www.instagram.com/p/{shortcode}/&omitscript=true

This returns thumbnail_url — useful for single images, but only gives the first frame for carousels.

Headers to include on all requests:

User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36
Accept-Language: en-US,en;q=0.9
Accept: text/html,application/xhtml+xml,application/json

Step 3 — Extract CDN image URLs

From the fetched data, extract all high-resolution CDN URLs. Instagram CDN URLs follow these patterns:

https://scontent.cdninstagram.com/v/...jpg?...
https://scontent-lax3-1.cdninstagram.com/v/...jpg?...
https://instagram.fXXX1-1.fbcdn.net/v/...jpg?...

For single image posts:

• Extract the single display_url or the largest display_resources entry (pick the one with the highest config_width).

For carousel posts:

• Look for edge_sidecar_to_children.edges[] in the JSON. Each edge has its own node.display_url and node.display_resources[].
• Iterate all edges in order. This determines slide numbering.
• Pick the highest-resolution variant from each slide's display_resources array.

For Reels:

• The cover image is extractable the same way as a single image.
• The video file itself requires a third-party tool (see Bonus section).

If JSON extraction fails, fall back to scraping <meta property="og:image"> tags from the page HTML — this gives at least one image URL (the first slide or only image).

Step 4 — Sanitise folder name

Build the folder name from the post caption:

1. Take the first 40 characters of the caption.
2. Strip all characters that are not alphanumeric, spaces, or hyphens.
3. Replace spaces and hyphens with underscores.
4. Lowercase the result.
5. Strip leading/trailing underscores.
6. If the result is empty (e.g. caption was all emoji), use the post shortcode instead.

import re

def sanitise_folder_name(caption: str, shortcode: str) -> str:
    truncated = caption[:40]
    cleaned = re.sub(r'[^a-zA-Z0-9 \-]', '', truncated)
    underscored = re.sub(r'[\s\-]+', '_', cleaned).strip('_').lower()
    return underscored if underscored else shortcode

Step 5 — Create output folder structure

import os

base_dir = "./instagram-downloads"
folder_name = sanitise_folder_name(caption, shortcode)
post_dir = os.path.join(base_dir, folder_name)
os.makedirs(post_dir, exist_ok=True)

If a folder with that name already exists (e.g. running the same URL twice), append the shortcode to avoid collision: folder_name_SHORTCODE.

Step 6 — Download each image file

For each CDN URL, download the file with a streaming GET request:

import requests

def download_file(url: str, dest_path: str) -> bool:
    headers = {
        "User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36",
        "Referer": "https://www.instagram.com/",
    }
    response = requests.get(url, headers=headers, stream=True, timeout=30)
    response.raise_for_status()
    with open(dest_path, "wb") as f:
        for chunk in response.iter_content(chunk_size=8192):
            f.write(chunk)
    return True

Name files:

• Single image: image.jpg
• Carousel slides: slide_01.jpg, slide_02.jpg, ... (zero-padded to 2 digits, or 3 digits if >99 slides)

Detect file format from the Content-Type header or URL extension. Instagram ser