I built this because I kept mass-deleting draft tweets that got no engagement. Wanted to know before posting whether something would actually reach people.
It scores against 36 rules derived from X's open-sourced algorithm (twitter/the-algorithm on GitHub). The part I find most useful is the what-if scenarios — remove a link (+52%), add an image (+38%), post at peak time (+25%).
Everything runs client-side in the browser. No account needed. AI features (slop detection, auto-optimize) are optional and use your own API key.
Happy to answer questions about the scoring model.
aytuncyildizli•1h ago
It scores against 36 rules derived from X's open-sourced algorithm (twitter/the-algorithm on GitHub). The part I find most useful is the what-if scenarios — remove a link (+52%), add an image (+38%), post at peak time (+25%).
Everything runs client-side in the browser. No account needed. AI features (slop detection, auto-optimize) are optional and use your own API key.
Happy to answer questions about the scoring model.