YAL's social media downloader

Quick links: documentation

Preparations

You'll need to have a vague understanding of what a "command prompt" or a "terminal" is.

1. Install node.js.
   Just about any semi-recent version should work.
2. Install CURL if you don't have it yet.
   CURL is included with most operating systems these days, see if running curl --version in a command prompt/terminal finds the app.
3. Some image-related features (like resizing or thumbnail generation) use ImageMagick.
   You might install it if you intend to use those or just in case.
4. Download this repository.
   That's ❮❯ Code ➜ Download ZIP, or "clone" it if you know Git
5. If you have downloaded a ZIP, extract it somewhere.

How to use

At its simplest, the tool is used like so (from a Command Prompt/terminal in bin directory):

node yal-dl.js https://bsky.app/profile/yellowafterlife.bsky.social/post/3lwh6s7axvc2o

and the tool will output something like

bsky=yellowafterlife.bsky.social=3lwh6s7axvc2o.jpg
bsky=yellowafterlife.bsky.social=3lwh6s7axvc2o=1.jpg

which are names of the images it just downloaded.

If you want to download images from several posts at once, you can supply additional argument(s):

node yal-dl.js https://bsky.app/profile/yellowafterlife.bsky.social/post/3lwh6s7axvc2o https://bsky.app/profile/yellowafterlife.bsky.social/post/3lwk6evrnec2o

and the output will look like so:

bsky=yellowafterlife.bsky.social=3lwh6s7axvc2o.jpg
bsky=yellowafterlife.bsky.social=3lwh6s7axvc2o=1.jpg

bsky=yellowafterlife.bsky.social=3lwk6evrnec2o.jpg

(an empty line separates media from different posts)

The tool is single-file so you can copy yal-dl.js wherever you need it, or even make it globally accessible.

Check out the docs for more parameters - you can have the tool take input from files, or output to a file, or even generate Markdown for a little gallery like this!

Gallery breakdown

links.txt:

Staircase & greenery
https://x.com/jhsuart/status/1428439445246672898

Lighthouse
https://x.com/LukeTscharke/status/1456391300907896832

Water ceiling
https://x.com/pasta_loving/status/1923581169339335071

...

Command:

node yal-dl.js --prefix img/ --in links.txt --out out.md --thumb 240x400

Markdown structure:

<style>
.columns {
	column-count: auto;
	column-width: 160px;
}
.columns > *:not(h2) {
	break-before: avoid;
	break-inside: avoid;
}

</style>
<div class="columns">

## [Staircase & greenery](https://x.com/jhsuart/status/1428439445246672898)
[![](img/twitter=jhsuart=1428439445246672898=th.jpg)](img/twitter=jhsuart=1428439445246672898.jpg)

...

</div>

Support

Website	S.?	Notes
Bluesky	✔	Uses VixBluesky to fetch URLs for videos and/or logged-in-only accounts.
Twitter	✔	Uses vxTwitter to fetch all metadata.
Tumblr	⚠	OG only contains the first image, consider using fxTumblr for multi-media posts.
Mastodon	✔	State-of-art OpenGraph support.
WAFRN	⚠	Off-network (e.g. bsky) posts may contain malformed image URLs¹.
YouTube	❌	OG only contains video thumbnails, consider using yt-dlp/etc.

Other websites are supported if they:

1. Include their media content in OpenGraph tags
2. Have posts distinguished by the URL (e.g. /user/cool-post) rather than HTTP GET parameters (e.g. /posts?id=101).

The bare minimum OpenGraph implementation has long been "first image/thumbnail and a little text" so you can typically count on that much.

Findings

This sounds like a one-evening project and it really should be, but turns out that everyone likes when other people implement OpenGraph, but do not like doing so themselves:

• Twitter no longer serves any OpenGraph content, so third-party services are the primary way to acquire media file URLs.
• Have you seen those mostly-functional Bluesky embeds in Discord? Those aren't OG.
  They also ignore the logged-in-only rules, as if that has ever stopped anyone.
• Bluesky can also serve you WEBPs or slightly lossy PNGs if you ask nicely.
  (change @jpeg at the end of a URL to @webp or @png)
• Isn't it kind of quirky that social media put usernames in og:title?
  Like if you supply a news article, image most likely relates to og:title, but for social media you need to read og:description instead. And how do you know which one it is?
• You can hand-write rules for other websites, but it is work (on average).

Considerations

Subject of expansion;

• Pinterest is okay if that works for you, though it has its limitations.
• Storing your mood boards in Discord/Telegram is one of the worse ideas, it's someone's else computers and they can kick you out whenever they feel like it (e.g. false-positives for spam).
• Pasting images into a folder works, but you might have a hard time figuring out where'd something come from if reverse image search comes up empty.
• Naming and organizing images/notes about them takes time and effort.

This tool makes it easier to download and organize images without opening each page yourself.

Building

You'll need Haxe installed.

haxe build.hxml

If all is well, bin/yal-dl.js will be updated.

TODOs

• Documentation
• Alt text
• An option to include post text in Markdown output mode?
• Add permalink to EXIF metadata (needs IM)
• An option to use <img> with width/height to avoid layouts reflows
• fxReddit, ddinstagram?