Home / Articles / The cell is the unit: positioning AgentiCCells around small, focused cells, each with an objective, a deliverable and a behaviour policy, each with a page that sells it, a free recipe and three ways to get help, a catalogue that grows one need at a time, in each industry's own words, and why we should run our own work in cells first
The cell is the unit: positioning AgentiCCells around small, focused cells, each with an objective, a deliverable and a behaviour policy, each with a page that sells it, a free recipe and three ways to get help, a catalogue that grows one need at a time, in each industry's own words, and why we should run our own work in cells first
By Dinis Cruz · 2026-11-16 · article v1.0.0 · newsroom v0.14.123 · agenticcellspositioningcellsproductgo-to-marketabparticle
Abstract: AgentiCCells needs a sharper idea at its centre, and I think it is the cell itself. A cell is the unit of what we sell: one focused job, with an objective, a deliverable and a behaviour policy, and as many parts as the job needs, a Claude session, a vault, a repository, a website, lanes and keys, a schedule, a connector, a person who owns it. A cell can be as simple as Claude replying to email, or as rich as a team keeping a website; every website is a cell. Now that the site factory creates repositories and changes AWS from a commit, we can put capabilities into cells that we never could before. So the selling has to be just as focused: every cell gets a page, with what it does, what you get, examples, use cases and its policy, and, on the side, four ways to start: the recipe for free, our agent helps you, we help you, or we set it up for you. The catalogue grows the simplest way: someone says "I need help with this", and we add a cell for it, often in the language of one industry, a clinic, a farm, a council. And we should live in cells ourselves: our team should think in cells, and I should move my own Claude work into cells, because that is the best test of whether they are easy to use, and the fastest way to find the next one.
In short
- The cell is the unit: one focused job, with an objective, a deliverable and a behaviour policy.
- As many parts as the job needs: a Claude session, a vault, a repository, a website, lanes, a schedule, a connector, a person.
- Every website is a cell, and so is Claude replying to email.
- Every cell gets a page that sells it: what it does, examples, use cases, its policy, and four ways to start.
- The first way is free: the recipe. Then our agent helps, we help, or we set it up.
- The catalogue grows one need at a time, often in one industry's own words.
- We live in cells first: the team, and my own Claude work.
Why the cell
AgentiCCells today explains itself by size: a ladder from one agent to a network of teams, the six setups of an agentic system, two modes of who runs the team (six setups of an agentic system). That is right, and it is how a buyer understands the price. But it is not how a buyer thinks about the problem. Nobody wakes up wanting "a few sessions with lanes and a vault". They want help replying to email, keeping a website, following up after an event.
So the cell itself should be at the centre: the unit of what we build, what we sell, and how we work. We need to think hard about what a cell looks like, how it is sold, what the customer is buying, how it is set up, and how it actually works day to day.
The anatomy of a cell
Every cell has three things it cannot do without:
- An objective: the one job it exists for, said in a sentence the customer would use.
- A deliverable: what you have at the end of each day, week or run: drafted replies, a live site, a debrief, a list of what waits on you.
- A behaviour policy: what it may reach, what it was asked to do, and what stands in its way (where assist becomes execute).
And then as many parts as its job needs: a Claude session to do the work; a vault for its memory, files and record (how a vault works); a repository and CI; a website on its own subdomain; lanes and keys to talk, sign and approve; a schedule of wakes; a connector to email, a calendar, Odoo or a CRM; and a person who owns it.
That range is the point. A cell can be as small as one Claude session replying to email under a policy, or as rich as a team keeping a website with a factory behind it (the site factory works). Every website is a cell. Now that the factory creates repositories and changes AWS from a commit, cells can carry capabilities that were too painful to set up before.
A page for every cell
Selling has to be as focused as the cell. Every cell gets its own page, built the same way:
- What it does, in one line, and what you get, concretely.
- Examples: real ones, from our own work and our customers', with links.
- Use cases: the situations where people reach for it.
- Its behaviour policy, in plain words: what it can reach, what it will never do.
And, on the side, four ways to start:
- Free: the recipe. The brief, the policy and the steps to run it yourself, with your own Claude. Most people should start here, and many will stay here, which is fine: it drives usage.
- Our agent helps you: a specialist agent reviews your setup and tells you what to fix (specialist agents for every team).
- We help you: two sessions with a person, on the ladder at £1,500.
- We set it up for you: in your accounts or ours, from £500 (atomic cells, end to end).
A catalogue that grows one need at a time
The catalogue grows the simplest possible way. Someone says: I need help with this. I need help sending these emails. I need help keeping this page current. And we add a cell for it. That is the workflow now: find the use case, write the cell, give it a page.
Some cells are general: reply to your inbox, follow up after an event, a weekly digest, a website you steer, translate your site, debrief your meetings, a CRM of your contacts, a policy for your agent, a vault for a project. Many will be specific to one industry, and that is good. A clinic's cell drafts patient letters in plain words for a clinician to approve; a farm's takes orders by email and keeps a page of what is in season; a council's answers the same questions every week, with the sources. The cell is the same underneath; its page, its examples and its words belong to that industry, and it connects to that industry's systems (integrate with whatever is on the other side).
Live in cells first
The best test of whether cells are easy to use is to use them ourselves, every day. Our team should start thinking in cells: what is this piece of work, what is its objective, its deliverable, its policy? And I should move my own Claude work into cells. Today much of it is sessions that come and go; as cells, each would come packaged with its brief, its policy, its vault and its record, and each would become an example on a page.
The same logic as the CRM applies: what we use ourselves gets better fastest, because the feedback loop is ours (a CRM for us and for our agents). Every cell we run becomes a page, a recipe and a proof.
What to build next
- A cell format: a folder with the objective, the deliverable, the behaviour policy, the recipe and the parts, the same for every cell.
- The page template for agenticcells.ai, with the four ways to start, generated from that folder.
- The first ten cells, from what we already do: reply to your inbox, follow up after an event, a website you steer, debrief your meetings, draft bills in Odoo, a policy for your agent, a vault for a project, a weekly digest, a page per place, CI behaviour policies.
- Our own work in cells: each of us moves one recurring piece of work into a cell this week.
- A request form: "I need help with...", into an agent's lane, which proposes the next cell.
What this does not claim
- The page and catalogue are mock-ups; the cells exist today as pieces of our work, not yet as a catalogue.
- Prices follow the AgentiCCells ladder, which its site calls first proposals.
- The industry cells are examples, not offers with customers yet.
Where this comes from
My own thinking, 16 November 2026, and the current agenticcells.ai and its cells page. Earlier on this site: the intersection, six setups of an agentic system, atomic cells, end to end, the site factory works, specialist agents for every team, integrate with whatever is on the other side, a CRM for us and for our agents, where assist becomes execute, how a vault works.
Peer review Not reviewed yet. Review it
Threads
Builds on
- The intersection: SGit, SHarness, Agent Behaviour Policies, RiskMandate and AgentiCCells as one stack of separate companies, now with real customers, real usage and real examples, one ladder of offers on every site, and a plan to align them all SGit, SHarness, ABP, RiskMandate and AgentiCCells as one stack of separate companies, with real usage and one ladder of offers on every site.
- Six setups of an agentic system: from a single task to one agent, sub-agents, several sessions, a conductor and its team, and conductors of conductors, with the altitudes of a run inside each and command and control growing at every step Six setups of an agentic system, from a task to conductors of conductors, with the altitudes of a run inside each: make the dangerous thing unreachable.
- Where assist becomes execute: tiering AI agents by what they can reach and whether it can be undone, how Agent Behaviour Policies drew that line for more than a hundred agents at Web Summit, and the rule that failed because it was only a rule Where assist becomes execute: tier AI agents by reach and undo class, enforce each tier with a boundary, and learn from the rule that failed.
- How a vault works: an encrypted store that acts like a database, where a server that is fully compromised leaks nothing, one mutable file lets every other object be cached for ever, and a web page can open a file in four requests or run a whole app in a sandbox How an sgit vault acts like a database on a server that cannot read: one key-derived mutable file, everything else immutable.
- The site factory works: three AgentiCCells sites made by hand, then one made by a commit in two minutes and thirty-nine seconds, live over https with a first version written by Claude Code, and the long list of what it took to make something that hard look easy Three AgentiCCells sites by hand, then one by a commit: live over https in 87 seconds, a first version by Claude Code at 159.
- Specialist agents for every team: long-lived agents that keep one skill, design, UX, code, security, behaviour policies or writing, with a person who owns it, so that new teams start from our best work instead of from scratch, and why that makes expertise that never scaled into something you can hire, even from another company Specialist agents for design, UX, code, security, ABP and writing serve every team, so new ones start from our best work. Expertise you can hire.
- Atomic cells, end to end: the £500 website as AgentiCCells' first offering, an order on a queue, an agentic team that builds a public site, a customer who steers it from their browser with a key, and a handover, with no daily work for the founder AgentiCCells' first offering: a £500 website as an atomic cell, steered by the customer from their browser with a key, handed over to their GitHub.
- Integrate with whatever is on the other side: why we never start by asking a customer to move, how we build around the systems they already run, from Odoo 12 to Salesforce, and how, over time, either their system gets better and we retire our code, or ours becomes the better place We never start by asking a customer to move. We integrate with what they run, build around it, and either retire our code or become the better place.
- A CRM for us and for our agents: why the next stage is a CRM of our own, built on vaults with an iOS app, why we will build it by using it first for our own team and our own fleet of agents, how the open-source and agentic CRMs compare, and the two teams it needs, one that builds it and one that uses it Our next stage is a CRM of our own, on vaults with an iOS app, for us and our agents. Twenty leads the field; agents propose, people send.
Continued by
- Show me the evidence: websites for people looking for a job, built by an agentic cell, with seven pages that prove instead of list, a page made for each employer, a pack for recruiters, eight personas to fork, a ladder from a free recipe to a recruiter's cell, and the brief for the team that builds it Websites for people looking for a job that show evidence instead of lists: seven pages, a page per employer, a recruiter pack, eight personas to fork.
- Fork it: making forking the way people adopt our cells and template sites, as GitHub made it the way people adopt code, with five ways to make a cell yours, what a fork-ready repository needs, paths from one repository to the next, a hub that lets usage decide what we build, and the brief for the team Forking as the way people adopt our cells: fork a template, point Claude at it, and it is yours. Five ways, fork-ready repositories, and a hub.
Article No. 436 · start a reading list with it
Reviews
No reviews yet.