A Content Agent That Builds Itself Around Your Business.
One interview installs it. It drafts on schedule, you review and feed back results, and every run gets sharper.
$997 one purchase · runs on your own Claude plan · no publish integration, you do the publishing
One Interview. It Builds All Of This.
Setup is a conversation in plain questions, about an hour. By the end, a working folder on your machine holds everything below, written for your business, in files you can open and read.
Its job
A written job description you sign. It checks the signature before every run.
Your business context
Your offers, customers, and positioning, drawn out of the interview and encoded.
Your voice and audience
Built from your own writing samples, with the platforms you publish on and who reads them.
Its memory
Plain files you own, where every correction and lesson lands permanently.
Its schedule
The runs that fire without you, set in the interview, changed whenever you like.
Its guardrails
Your boundaries, confirmed with you and written down. It works inside them or it stops.
Its results record
A ledger of what you published and what happened, feeding what it makes next.
Its work journal
Every action it takes, written down in plain language you can audit later.
What A Morning Delivery Looks Like.
You open the folder with your coffee. The day's brief, drafts in your voice, a note on what is waiting for you. Every decision you make on them becomes part of how it thinks. The real folder, as delivered, is on the receipts page.
├─ daily-brief-2026-07-29.md · the day's content brief
├─ draft-facebook.md · in the owner's voice, inside the brief
├─ draft-linkedin.md · in the owner's voice, inside the brief
├─ needs-attention.md · what is waiting on the owner, marked READY
└─ journal.md · every action, on the record
It Gets Better Because You Stay In It.
The daily routine takes about five minutes, and it is where the compounding comes from. The engine's own guide puts it plainly: feed it results and it learns your audience. Starve it and it stays generic.
1 · READ
Open the morning folder. The brief and drafts are waiting.
2 · SHIP
Post what you approve, edit what is close, reject what is off.
3 · ANSWER
Two quick results questions: what went out, and how it did.
4 · IT LEARNS
Your answers and corrections are written into memory, permanently.
5 · IT ASKS
Any change of strategy comes back to you for approval first.
Your judgment is the asset: six months of your corrections builds an agent nobody else can buy, because they would need six months of you.
The Content Intelligence Center.
The engine ships with the Content Intelligence Center, a dashboard for the visual types, and it is completely optional: the folder and your morning chat are the full experience without it. It is a window into the agent's files; the files stay the product. Out of the box it is a talking reporting dashboard, reading you the state of play. Pair it with the free Claude Code CLI, the one slightly technical option in the whole product, and it becomes fully interactive: you ask, it answers. Here it is in motion.
What A Run Actually Does.
When a scheduled run fires, this is the whole job, in order, on the record.
It checks its signed job description is unchanged and reads your feedback inbox, so anything you noted yesterday shapes today.
Research comes next: your market read against your themes, what moved and why it matters, sources kept.
It drafts for the platforms due today, in your voice, weighted by what your results ledger says has landed before.
Its own reviewers check every draft against your writing rules before you ever see it, and rewrite what fails.
The finished brief lands in your folder, every action journalled, and the run closes down until the next one.
AND THE THINGS IT CANNOT DO
It ships with no publish integration at all: no account connections, nothing to send with, nothing to spend with, no way to contact anyone. Those are structural facts about what was built. Every draft waits in your folder, and putting work into the world stays your act alone.
Your First Week.
DAY 1
It installs itself and interviews you. About an hour, all conversation.
DAY 2
First scheduled run. A brief and first drafts wait in your review folder.
DAYS 3-6
You work through the drafts. Every decision becomes permanent instruction.
DAY 7
Keep the five-minute routine and drafts arrive noticeably closer to shippable. That gap keeps closing.
Yours, On Plain Terms.
THE INSIDE OF THE BOX
Every file it writes for your business is plain text you can open. Nothing hides behind a dashboard.
├─ brief.md · the job, signed by you
├─ context/ · your business, encoded
├─ memory/ · BIMRI, plain files you own
├─ strategy.md · changes proposed to you first
├─ feedback-inbox.md · your notes, read every run
├─ artifacts/ · the finished work
├─ journal.md · every action, on the record
└─ security-log.md · every boundary event, noted
BEFORE YOU BUY · THE PLAIN FACTS
PRICE · $997, one purchase
LICENCE · perpetual, one business: yours. Your inputs, outputs, memory and workspace are yours. Reselling or redistributing the files is the one thing it never allows
RUNS ON · your own Claude plan, about two scheduled runs a day
PLATFORMS · Claude Code and Claude Cowork, documented and supported
RUNTIME · scheduled runs fire while the app is open on your machine; on the Claude Code command line, schedules renew with one word about weekly
SKILL LEVEL · if you can install an app and hold a conversation, you can run this
PUBLISH AUTHORITY · none. You review, you ship
DELIVERY · install package and guide, by email, minutes after purchase
SUPPORT · a human by email, answered within 3 business days
UPDATES · your version is yours forever. Major new versions are optional paid upgrades
The Straight Answers.
What does it run on?
Claude Code and Claude Cowork are the documented paths; those two are what we test and support today. Scheduled runs fire while the app is open. If your machine was off, the run fires when you next open it, and the morning message tells you what it missed. One quirk worth knowing upfront: on the Claude Code command line specifically, repeat schedules wear out after about a week, and saying "morning" renews them. The guide covers it.
What does it cost to run?
Scheduled runs consume your own Claude plan's usage, at roughly two runs a day. Setup states this plainly before you schedule anything. The engine itself is one purchase, licensed to your business forever.
What does the interview need from me?
A spare hour and honest answers. It asks plain questions about your business, your audience, and where you publish. Three to five samples of your own writing sharpen its ear on day one, and if your folder already holds business documents, it reads those first and asks less.
Can I see inside it?
Yes. Every file it writes for your business is readable and yours: the identity, the rules, the memory, the strategy file. All of it sits as plain text in your workspace, where a dashboard can never lock you out.
Will it post things without me?
No. There is no publish integration: no account connections, nothing to send with. Drafts sit in the review folder until you move them yourself.
What happens when my business changes?
Tell it. A line in the morning chat or a note in the feedback inbox lands in its files and holds from the next run. Bigger shifts in direction come back to you as a proposed strategy change, so nothing moves without your yes.
What if it drifts or gets something wrong?
Correct it once and the correction is written into memory, so it stays fixed. The journal shows you every action it took and why, and it is built to stop and ask when its own checks fail rather than improvise.
What if installation stalls?
The install is a guided conversation, and the guide covers the usual snags. If you get stuck anyway, email support and a human works it through with you, within three business days and usually much faster.
Are updates included?
Your purchase is version-stamped: the version you buy is yours to keep and run forever. Major new versions are optional paid upgrades, announced to the waitlist first.
Why not just use a chat window for this?
Chat is brilliant for thinking out loud, and we still use it that way. This system is built for operations: a schedule that fires without you, files it reads before every run, corrections that stay made, and results that shape the next output. Keep feeding it results and the difference shows by week two: drafts arriving already shaped by everything you taught it in week one.
The First Of The Line.
The Content Intelligence Engine is the first Agent in a Box. Client acquisition, operations, and agency project management are next, on the same spine.