The Sent2X playbook
How the product fits, how to grow, and what each plan can do
First the growth loop and the operating points, then the quota rulebook and plan capacity. Opening this page spends nothing.
Architecture & growth
Chapter 1
How Sent2X fits together
One loop, not six separate apps. Connect X, keep a voice, find rising posts and people who would buy, reply in that voice, publish on a cadence, and read Home to see what pulled followers.
| # | Piece | What it does |
|---|---|---|
| 1 | X connect + extension | Sign in with X so posting is authorized. The browser extension docks a panel beside x.com and puts three buttons in every reply box: automations, the AI pen, and voice input (any language → English). Translate is a module on that automations list. The panel is an iframe of /embed; full pages like this one open in a new tab so they are not blocked by X's frame rules. |
| 2 | Voice & experiences | We seed a voice profile and an experiences document when you join (on us). Re-learn / Generate is once per local day, also on us. Replies, Remix, and Planner read these so drafts sound like you. Your own edits are kept. |
| 3 | AI Discovery | Topic channels for your niche. A search counts when it serves you a batch of active cards — from a fresh fetch or a preheated pool. Reloading the page or iframe that only re-reads cards already in the pool is free. Neighbour prefetch is on us. Each counted search tops a channel toward 15 cards; ranking usually keeps about 20 usable. |
| 4 | Buyers | Paste a product homepage. We search X, group accounts into watch lists (looking for a tool, competitor complaints, indie hackers, tech creators, side-project founders), and seed reply templates. Import the list, then open Discover to see their new posts. Each successful analysis spends one Buyers turn — Free 2, Pro 20, Max 50 a month. Planner is a separate pool. |
| 5 | AI replies + Remix | Blue-pen on x.com or on a Discover card, voice input, custom quick-replies that call the model, extension Translate, Remix drafts, and Planner-plan translation share one pool. Let's connect templates do not count. Open & draft opens the composer from a Discover card without spending a slot. Type or speak your own angle on a real thread — that is the growth move. |
| 6 | Replies inbox | Mentions land in one place with an AI draft waiting. This is inbound: people already talked to you. Turning those into real replies compounds more than spraying more cold comments. |
| 7 | Planner, queue, agents | Each Planner message you send counts as one, including follow-ups in the same chat. Import lands in the queue. Overnight draft agents (Pro 1 / Max 5) draft overnight in your voice and wait for approval. Extra Run now is separate (Pro 3 / Max 10 per day, shared). A post slot is spent only when it actually publishes to X. |
Home (the dashboard, or the panel's Home tab) is the pulse: followers, impressions, engagement, next queue slots. Analytics warm up after X is connected — give it a couple of days.
Chapter 2
How to grow fast on X
Five points, not a 20-item list. They match how Sent2X actually meters and how X ranks early, substantive replies.
01
Show up every day
A daily loop beats a weekend binge. Discovery and replies still move the account on days you do not post. A missed day just means the loop did not run.
02
Reply early, on threads you would actually join
X pays for early, useful replies on rising posts. Open Discovery daily, prefer posts that are still warming up, hit View to open the post. After a Buyers run, open Discover on that product to reply to those accounts' new posts. Spend the reply pool on threads in your niche — not on every card the channel can show. 150 replies/day is a ceiling, not a target.
03
Write in your voice, not the template pool
Type a half-formed thought before you hit the blue pen. Connect-post templates are free because they do not call the model — they also do not grow an account. Voice and experiences are on us; keep them current so drafts do not drift into generic AI.
04
Do not burn Discovery on empty refreshes
Opening a channel that already has cards does not count. Warming neighbouring channels is on us. A search is spent when the pool is empty and we actually fetch. Skip or post to free a slot; wait until a channel is below 15 active cards before you force a refill.
05
Close the loop: inbox + a posting cadence
Answer mentions from the Replies inbox — a reply that earns a reply back is worth more than another cold comment. Keep the queue fed with Planner or Remix so Home is not empty. Quotas buy volume of this loop; they do not replace showing up.
Numbers and meters: see Usage & billing. Recorded walkthroughs of each engine: How to Use.
Usage & billing
Chapter 1
How quotas are counted
Each meter has its own window. The table is the product rule — not a guess.
| Quota | What increments it | Window | What does not count |
|---|---|---|---|
| Posts published to X | One slot when a post actually goes out: scheduled fire, Publish now, or post immediately. Remix-agent posts and hand-written posts share this ceiling. A failed publish is refunded. | Free: calendar month in your timezone. Pro / Max: your Stripe billing period (the weekday you subscribed), not the calendar month. No daily post cap. | Drafts. Queued or scheduled posts that have not published yet. Deleting a draft. Opening this page. |
| AI replies + Remix | Each time the model writes or translates: the blue-pen button (when it actually calls the model), voice input (speech → English), a custom quick-reply that calls the model, Translate in the extension, a manual Remix from Discover, and translating a Planner post plan. Remix and replies share one pool. Agent drafts do not spend this pool. | Your timezone, not UTC. Free: 5 per day and 30 per month. Pro: 150 per day. Max: 400 per day. Paid plans have no monthly reply cap. | Let's connect templates (no model call). The blue pen always calls the model and counts, including on connect posts. |
| AI Discovery searches | Counted when a search serves you a batch of active cards — including cards already waiting in a preheated pool. Fetching new posts from X is not required for the count. Each counted search tops a channel toward 15 cards on screen; ranking usually keeps about 20 usable posts — you see 15 first, then the rest as you skip or reply. | Free: 30 searches per calendar month in your timezone. Pro: 20 per local day. Max: 30 per local day. | Refreshing a page or iframe that only re-reads cards already in the pool. A search that comes back empty. Warming neighbouring channels (we pay). Skipping, viewing, or posting a card. Cards themselves are not a separate meter — the served search batch is. |
| AI Planner messages | Each message you send in AI Planner counts as one — a new chat's first message and follow-ups in the same chat. Counted when the model actually runs. | Calendar month in your timezone. Free: 30. Pro: 150. Max: 400. | Translating a finished plan — that spends an AI reply, not a Planner turn. A send that fails before a reply starts. Buyer analyses use a separate monthly pool. |
| Buyer analyses | Each successful Buyers run counts as one: the first homepage analyze and each follow-up that actually runs the model (re-search or rewrite). Failed fetches (site down, 404, thin page) do not count. | Calendar month in your timezone. Free: 2. Pro: 20. Max: 50. | Importing accounts, opening Discover for a Buyers list, filling a reply template, or a send that fails before the model starts. Planner messages are a different meter. |
| Overnight draft agents | How many agents you may keep: Free 0, Pro 1, Max 5. An agent that is on runs once a night without spending extra Run now. Each post that actually publishes spends one post slot. | Agent count is a standing cap, not monthly. Extra Run now is a separate daily meter. Publishes use the post window. | An idle agent. Creating or editing an agent. Manual Remix still works on Free (it spends the reply + Remix pool, not an agent slot). |
| Extra Run now | Each extra Run now click on an agent counts as one. Shared across every agent on the account. The nightly auto run does not spend this. | Your timezone, not UTC. Free: 0. Pro: 3 per day. Max: 10 per day. | The nightly run when an agent is on. Creating or editing an agent. Manual Remix from Discover (that spends the reply + Remix pool). |
| Voice profile & experiences | We build both once when you join (on us). After that, Re-learn or Generate once per local day — also on us. Each run merges new facts; your own edits are kept. | Manual rebuild: once per local day. Not a paid meter. | Does not spend replies, Discovery, Planner, Buyers, or posts. |
Limits above match the live product caps. Discovery 20/day ≈ 300 cards on screen / 400 usable if you work through that day; 30/day ≈ 450 / 600. Those are Discovery cards (skip, blue-pen, Remix), not your monthly post allowance (Free 10, Pro 500, Max 750). Buyers is its own monthly pool (Free 2, Pro 20, Max 50). A blue-pen, voice input, translate, or Remix still spends an AI reply. Open & draft and the Replies inbox themselves do not.
Chapter 2
What each plan can do in a month
These are the hard ceilings for each meter. Opening this page spends nothing. What you actually get on X still depends on niche, account age, and whether you use the quota.
Monthly capacity
| What you can do | Free | Pro · $39/mo | Max · $59/mo |
|---|---|---|---|
| Posts published | 10 / month | 500 / month | 750 / month |
| AI replies + Remix | 5 / day, 30 / month | 150 / day, no monthly cap | 400 / day, no monthly cap |
| Discovery searches | 30 / month | 20 / day | 30 / day |
| Cards if you work through searches | ~450–600 / month | ~300–400 per full day | ~450–600 per full day |
| Planner messages | 30 / month | 150 / month | 400 / month |
| Buyer analyses | 2 / month | 20 / month | 50 / month |
| Overnight draft agents | None (manual Remix only) | 1 | 5 |
| Extra Run now | None | 3 / day, shared | 10 / day, shared |
| Reply box + inbox + Open & draft | Included — voice & translate spend the reply pool | Included — same tools, larger reply pool | Included — same tools, largest reply pool |