Once you've pasted in a tweet draft, you get a score with suggested improvements for the hook, specificity, answerability, and bait phrases for your draft, as well as a compare feature for two drafts. It's free and doesn't require an account.
The checklist is based on open sourced ranking code from X themselves (xai-org/x-algorithm), some research studies, and some published data from Buffer. Claude helped me comb through these resources and I turned them into these rules. I haven't tested the score against tweet performance, so I can't claim it has any bearing on virality, but I have applied the rules to drafts to check them.
It's free and doesn't require an account; the big box on the right is an ad for my other project, TweetBrain, a tweet drafting tool that's in development. It's completely separate from this one, and I'm doing this mostly as a self-reminder to promote it, since it's my thing and I should tell people about it myself.
Let me know what rules seem wrong to you, or if something like this is useful at all.