See how a draft will land before you post.
Yubook is a publishing coach for people who post on X. It scores a draft against X's published ranking signals before anything goes out, and it finds the posts where a reply would put you in front of someone else's readers. It does not trade follows, and it does not post to X for you.
The problem
A small account can publish every day and still not get seen. The usual advice is to post more, or to swap follows until the number goes up. Follow-for-follow is fast for a week. Those people never read you, and you cannot get rid of them.
Most “algorithm scores” are a black box. They give you a number and no sentence to change. Yubook uses the ranking weights X has published, splits a draft into signals, and tells you what the writing is asking people to do — reply, share, click, or mute.
How it works
You keep the sentence. Threads can go out from Yubook when you press publish. X stays in your own composer.
Draft signal analysis
Score a draft before it is posted. The result is not one opaque number. You see which lines pull for a reply, a share, a click, or negative feedback, and you can rewrite the sentence and watch the reasoning update.
Forecast range
Predicted reach is a calibrated interval. It is not an official X score and it does not guarantee views or followers.
Reply target queue
When you open the app, Yubook polls recent posts (10-minute cooldown, last 60 minutes) and ranks where a reply would still land in front of someone else's readers.
Watched accounts
The main pool comes from your following list. A secondary pass finds accounts by category. News and broadcast accounts are excluded. You can add more watch slots if you need them.
Seven-day growth plan and calendar
A click-through onboarding (7 steps) builds a week of reply and post targets. You confirm days. The calendar follows that plan.
Follower tracking
Your own follower count is collected once a day. Looking up your own account is unmetered, so this stays free.
AI drafts, replies, and rewrites
When you want a starting sentence or a reply, Yubook writes in the language you chose. You edit before anything is published.
X + Threads
Connect Threads and publish there when you choose. Yubook does not auto-post to X.
Free tools (no account)
X account status check — yubook.app/x-check
See whether search is still showing an account's posts. Official X API only. What it can see and what it cannot are listed separately. It will not invent a “no ghost ban” all-clear.
Best posting hours — yubook.app/best-time
Not a generic “post at 9am.” Median views from that account's own posts, in your timezone.
What Yubook will not do
4.8
6 ratings
AI video with native sound, paid per clip. Nothing stored.
A calmer Mac menu bar, one click away.
Enterprise Software. Intelligent Automation. Global Impact.
Comments
26Marketplace where AI agents bid on verifiable research jobs. Informational alpha.
Splitting the score into per-sentence signals instead of one opaque number is the right call - 'this line pulls for a mute' is something you can actually act on. The reply queue ranking by freshness + audience + competition is a smart constraint too. Does the forecast range calibrate differently for small accounts, where published ranking weights behave differently?
Building PZERO, saving you money - pzero.studio
Yubook splits a draft into the ranking signals X has published and tells writers whether the post is asking people to reply, share, click, or mute. If you need to make room in an AI budget, PZERO is at pzero.studio.
这个感觉还挺需要的，不知道效果如何？回头体验着试试
这个感觉还挺需要的，不知道效果如何？回头体验着试试
Showing the forecast as a range and breaking a draft into reply/share/click/mute signals is much more honest than the single 'virality score' most tools give. The line-level feedback is what makes it actionable - we run a small brand account and the usual failure is that a post asks for nothing, so nobody replies. Two questions: since X's open-sourced ranking weights are from 2023, how do you keep the model from drifting as the real algorithm changes - do you calibrate against actual outcomes of users' posts? And does the reply queue let you filter by topic/keyword, so a niche account (say ecommerce ads) only sees threads where a reply is on-topic rather than just fresh?
The draft preview idea is useful. Being able to see how a post will land before publishing would save a lot of guesswork.
AI-powered productivity app that turns saved content into action
The fact that Yubook won't post to X for you is what stands out most here — most "growth" tools in this space quietly turn into automation the moment you connect your account. Scoring a draft against the actual published ranking signals (reply/share/click/mute) instead of one opaque virality number is genuinely more useful feedback for a writer. Curious how the reply queue handles accounts that post in bursts vs. steadily — does freshness ranking favor the steady posters, or does a burst temporarily crowd out the queue?
The focus on actionable signals—what to reply, share, click, or mute—makes this more useful than a simple score. A before-and-after example would be a great way to show how the coaching changes a draft.
Breaking down drafts by specific engagement signals rather than a generic virality score is brilliant. Saves so much guesswork before hitting publish - definitely giving this a spin!
Hadaa Inc
Does it connect via API?
The onboarding copy is unusually clear about who the product is for, and the pricing table makes the free vs paid tradeoff obvious without hiding the backlink requirement.
Scoring a draft before it goes out is the step most writing tools skip: they help you write, then leave distribution to luck. One thing I'd want to see is how the score treats replies differently from root posts — the signals that get a reply seen inside someone else's thread feel different from the ones that rank a standalone post. Keeping it a coach instead of auto-posting is the right scope call.
Will this work for Blog posts as well?
I like the per-sentence signal approach because it makes the feedback actionable instead of reducing a post to one opaque score. The combination of reply targeting and the 7-day growth plan also seems useful. How are you measuring whether the forecast range stays calibrated across smaller accounts?
Founder of LaunchOn.it
looks good, will give it a try!
Building — 💻 Framer templates & components: https://www.framer.com/@avirallakhanpaul/ 🎥 Website Scroll Video: https://websitescrollvideo.com/
That is something genuinely useful!
solo developer
The "no sentence to change" line nails why most algorithm scores are useless — a number without a fix is just anxiety. Scoring the draft into what it's actually asking people to do (reply/share/click/mute) is the part that's genuinely actionable. And refusing follow-swaps is the right call; borrowed followers never read you anyway.
momo
Draft signal analysis Score a draft before it is posted. The result is not one opaque number. You see which lines pull for a reply, a share, a click, or negative feedback, and you can rewrite the sentence and watch the reasoning upd
IndieHacker
The decision to display a forecast range instead of a single promised view count is a refreshing way to handle the unpredictability of social algorithms. The reply target queue seems particularly useful for smaller accounts trying to find high-impact conversations without spending all day scrolling. How does the system determine the negative feedback or mute signals within a specific draft?
As someone who spends time building in public on X, a tool that scores drafts and acts as a publishing coach is brilliant. Definitely going to test how it helps optimize engagement. Awesome launch!