X Reply Agent
The posts where somebody is asking for exactly what you built go past while you are shipping. This agent reads them for you: every day it searches X, keeps the ones where you genuinely have an answer, and drafts the reply. You post it in the thread yourself, from your own account.
You saw it four hours too late
Somebody with 40,000 followers asked their timeline which tool to use, and four people answered before you saw it. That happens most days. The only way to catch it is to sit in X all afternoon, which is the opposite of building, and the reason most founders give up on replies about a fortnight before they would have started working.
What is reply marketing on X?
Reply marketing on X is answering other people's posts instead of writing your own. Somebody says they are fed up with the thing you replaced, or asks the room what they should use, and you answer them. Done properly it is the cheapest distribution on the platform, because you are borrowing an audience that has already stopped scrolling. Done badly it is the fastest way to get muted, because a reply that opens with your product name reads as an advert and everybody can tell.
The hard part is not the writing, it is the finding. There are a few posts a day where you have something true and useful to say, and they are buried under thousands where you do not. An X replies agent does that search for you: it reads the day's posts, keeps the ones where your product is the honest answer, and drafts the reply. Kwern's runs as a daily Loop and carries the post it is answering into the draft, with one sentence on why it thought that post was worth your time.
What it leaves you
One reply per post worth answering, each carrying the post it answers and the reason it was picked, and beside them the likes, reposts, follows and direct messages worth making, each carrying the post or the profile you make it from by hand. Your own posts come from the X posts Agent, on the same channel.
The skills it runs
A run may activate these and nothing else. Every one of them is open: read the file before you turn the agent on.
What an X reply agent costs elsewhere
These are the prices on those companies' own pricing pages on the day shown. Kwern publishes no audience figure for X, because we have not measured one and the ones on other people's pages carry no source.
Kwern against the alternatives
| What is compared | Kwern | AI CMO suites | Point tools | An agency | Doing it yourself |
|---|---|---|---|---|---|
| Replying without reading as an advert | Answers the question first and mentions your product only where the product is the answer; every draft says in one sentence why that post was worth answering. | Some reply automatically, which is how an account gets muted. | An alert tells you a keyword fired. What you do next is on you. | A junior replying from an account nobody recognises. | You, replying to the first thing you saw, which is rarely the best thing. |
| Finding the post worth answering | Searches the day's posts for the ones where your product genuinely belongs, and carries the post it answers into the draft so you can see what it read. | A keyword search ranked by engagement rather than by fit. | A keyword alert, which is every post, not the right one. | A weekly list, by which time the thread has moved on. | Whatever crossed your timeline while you were meant to be building. |
| Writing the reply | Drafted in your brand voice against the foundation document Kwern wrote about your product. | A template with the product name swapped in. | Nothing. An alert does not write. | Generic, because they do not know what your product does. | Good when you have the time, which is the problem. |
| Who says yes before it goes out | One feed. Approve, edit or reject each draft; nothing is published or spent until you do. | Usually a dashboard, sometimes autopilot with no stop. | Per tool, in whatever queue that tool keeps. | A shared document, and a reply a day later. | You wrote it, so you approved it. |
| What it costs a month | From $29 a month, every agent in every plan. | $20 to $1,000 a month, and $5,000 for a done-for-you tier. | One subscription per channel, each with its own seat. | A monthly retainer, agreed before anything ships. | Nothing, and it costs you the evenings. |
| Who pays for the model | Your own key. Across the 11 models Kwern vets, input runs $0.10 to $10 per million tokens, billed to you by the provider. | Credits you buy from them; the markup is never itemised. | Theirs, folded into the subscription. | Inside the retainer, at whatever margin they keep. | Your own key, at the provider's price. |
RiseMore, the cheapest
- www.risemore.app/pricing
MarketOwl, the dearest
- marketowl.ai/pricing
This Agent runs about 30 times a month. The tokens those runs spend are billed to you by your own provider: $0.10 per million in on gpt-6-luna, $10 on gpt-6-astra. Kwern never resells a token and never marks one up.
Three plans, every agent in each
Your model key, your accounts, your approval on every post. The plan decides how many products you run.
Questions about the X Reply Agent
Where to go next
Someone is asking for what you built today.
On Reddit, on X, in ChatGPT's answers. Kwern finds them every morning and drafts the reply in your voice; you decide what goes out.