Social Media Tools
Tools for making, checking and measuring what you post — all of them running in your browser. Nothing you upload, paste or type here is sent anywhere, there is no account, and none of it connects to a platform.
What this is not
There is no scraping here. Nothing reads comments, follower lists or anyone's analytics — you supply the file or the figures. Nothing posts on your behalf or asks you to log in to a platform, and nothing generates a fake screenshot, a fake message or an invented metric.
Where a tool depends on something a platform decides — an image size, a caption limit — the figure carries the date it was last checked, because those change without notice and a number quoted with confidence is worse than one quoted with a date.
Write it
Counting a caption properly, and finding what will break when you paste it.
Pictures for posting
Getting an image the right shape and size before a platform does it for you.
Before you post it
The things in a file that you did not mean to publish.
Measure it
Numbers that mean something, with the denominator stated.
Run a promotion
A draw anyone can check, and a link people can scan.
The business of it
Pricing work and knowing whether a campaign paid for itself.
Where the interesting problems were
A caption is not as long as your counter says. Ask how long a family emoji is and there are three defensible answers: one character to a reader, seven code points underneath, and eleven in the units JavaScript counts by default. Almost every counter reports the last one, so an emoji-heavy caption is described as hundreds of characters longer than it is.
Dividing an image into three is easy until the width is odd. A 1081-pixel picture cut into three gives 360.33 pixels a tile; round it and you lose a column of the photograph, round up and you repeat one. On a profile grid that shows as a seam. Working out where each boundary falls, rather than how wide each tile is, means the pieces cover the original exactly — and every cut is checked against that before anything is exported.
A draw can be biased without being wrong-looking. Taking a random number modulo the number of entries favours whoever is at the start of the list, every time, for as long as you keep using it. Discarding the values that fall in the uneven tail costs nothing and makes the draw genuinely uniform.