Back to blog

How to map your AI org chart before you build any AI employees

by Samantha ✦ 17 September 2026 ✦ 11 min read

The Incuv8or AI Framework - step 2 of 6: Map your org. Before this: build your AI brain. Next: build your COO.

We have all been there. We've all been chasing the AI agent hype, creating every single agent, downloading every single skill, and hoping that our systems and processes will get better.

But what nobody is telling you is that that is the worst way to actually implement AI employees into your business and into your life.

When I started playing with AI employees, I just started willy-nilly building them, creating skills, creating routines, and it got messy very quickly because I didn't really remember what I had. Half of them I didn't even need or use, because they were tweaked and targeted to someone else's business. I could easily have run off down a rabbit hole and created 197 agents when I'm only ever going to use about 10 of them.

So the first thing you actually need to do is stop downloading every single skill and trying to create all these different agents to do all these different things, and start with the boring stuff. The boring stuff is mapping your AI org chart.

How I actually work: one COO, a small team and a library of skills

I basically only ever communicate with one AI employee, my COO. I talk to it the most, I brain dump to it, and it hands the work out. It runs on Opus.

Underneath my COO, my business is split into departments, the same way a real company is. I have eight, because I run more than one business:

  • Business Operations - legal, finance, admin and the day-to-day running of the business
  • Clients - my freelance clients
  • Marketing - my content, social media and email
  • Sales - leads through to new clients
  • Product - the products I'm building
  • Education - Tech Queen Academy
  • Technology - my websites, integrations and security
  • Personal Life - my life admin, and the personal side of things

If you're a freelancer starting from scratch, you probably won't need all of those. The AI Org Mapper starts you with five - Business Ops, Products & Services, Client Services, Marketing and Personal - and you rename or drop them to fit your business.

Each department gets an AI employee when there's work that needs one, and each employee has a toolkit of skills. Skills are literally just job descriptions. A skill is a file that teaches the AI to do one process or SOP the way you do it in your business. My skills sit in one shared library, and each employee has an index of the skills it uses.

In marketing, for example, I have a Blog Manager, and my Scriptwriter has a skill for each format, like talking heads and B-roll captions. In products, I have more than one product, so each product gets its own employee, and they all pull from the same skills.

Each employee is a folder. Inside it is the employee's identity file, which says who it is, what it does, which Claude model it runs on and what it's allowed to touch, plus the index of its skills.

When a job gets its own employee

This is the part that keeps it lean. Anthropic's guidance is to start with one general agent and a library of skills, and only add more agents when a simpler setup falls short. So a job only gets its own employee when it needs one of four things:

  1. Its own context - a big, focused area of knowledge, like one product.
  2. Different permissions - it's allowed to touch things the others shouldn't.
  3. A different model - a cheaper one for simple work, or a stronger one for hard work.
  4. To run in parallel - it needs to work at the same time as the others.

If a job doesn't need any of those, it's a skill, not an employee. Most people only need a COO and three to five employees to start.

Anthropic explains the skills side of this in Equipping agents for the real world with Agent Skills, and their engineers make the case in their talk Don't Build Agents, Build Skills Instead.

The order you build it in: The Incuv8or AI Framework

  1. Build your AI brain. Your business, written down where AI can read it.
  2. Map your org. The AI Org Mapper turns your brain dump into a COO, departments and the skills you need, and saves your org chart in your AI brain. That's this post.
  3. Build your COO. The one employee you talk to. It knows your org chart.
  4. Add employees only when they pass the test. Your COO drafts each one with you, starting with your top five. Anything that fails the four-part test is a skill in your shared library.
  5. Hand out the work from your task manager. The same statuses on every list, a field for the employee, and a dispatcher that picks it up.
  6. Add routines where they're needed. Only for the jobs that run on a schedule.

Across every step: nothing goes live without your approval, and every correction becomes a rule your employee keeps.

The AI Org Mapper only maps your org chart. It doesn't build anything. It tells you what to build and in what order, and your COO drafts it with you.

What you'll have at the end

You install one skill, the AI Org Mapper, and it takes you through a brain dump session about how you run your business. When you map out your organisation, you're mapping out all of the different processes and SOPs in your business. Those become your skills, and the AI Org Mapper works out which departments and employees they belong to. At the end it gives you this.

MY AI ORG CHART

Guardrail: nothing goes live without my approval.

Me
└── COO - the one I talk to; hands out the work
    ├── BUSINESS OPS
    │   └── [Employee] - skills: [skill], [skill]
    ├── PRODUCTS & SERVICES
    │   └── [Employee per product] - skills: [skill], [skill]
    ├── CLIENT SERVICES
    │   └── [Employee] - skills: [skill]
    ├── MARKETING
    │   ├── [Employee] - skills: [skill], [skill]
    │   └── [Employee] - skills: [skill]
    └── PERSONAL
        └── [Employee] - skills: [skill]

WHY EACH EXTRA EMPLOYEE IS SEPARATE
- [Employee] - own context / permissions / model / parallel

BUILD THIS WEEKEND - the top five
1. COO
2. ...
3. ...
4. ...
5. ...

STAYS WITH ME (or a human)
- ...

Download the AI Org Mapper skill →

Your departments will be your own. Rename them, or drop the ones you don't need. Everyone's org chart is going to be different because everyone works differently. Some people are freelancers. Some people already have a team. The AI Org Mapper builds yours around how you actually run your business, not around mine.

Why the permissions matter

When we're thinking about how we structure the org chart, we need to think about permissions and guardrails.

In my setup, no AI employee publishes or changes anything live on its own, and that includes my COO. The guardrail is my approval. Nothing goes live until I approve it by setting a status on the task.

My Blog Manager is a good example. It writes the post and sends it back to me for review. It only publishes once I've set the task to "w/ agent to publish". Until I do that, it can't.

Where I'm heading next is proper permission levels, where my COO can act on the bigger things once I've approved them. That's why the AI Org Mapper asks you what has to stay with you or a human, so it's written on your org chart before you build anything.

What you need before you start

You don't need to connect any tools, and you don't build any skills in this session.

Step 1 - Install the AI Org Mapper

Unzip the download. Inside you'll find a folder called org-mapper with a file called SKILL.md in it.

  1. Open the Claude desktop app.
  2. Go to Customise.
  3. Then Skills.
  4. Then Add Skill.
  5. Upload the SKILL.md file from the org-mapper folder.

It's now available in every chat. If you want to know more about how skills work, it's in How to Build a Claude Skill.

What you'll see: AI Org Mapper in your list of skills.

Step 2 - Start the session

Click the Claude Code tab and start a new chat. Click the + at the bottom of the chat and choose your AI brain folder, so the AI Org Mapper can read your business and save your org chart into it.

Then type this.

Hey Claude, I want to use /org-mapper to map out my AI org chart.

It starts by telling you what it's already read in your AI brain, so you know what it knows.

Step 3 - Answer the questions

The AI Org Mapper takes you through a process of understanding what you do in your business, how you run it, and all of the different things you need. It asks one question at a time, and it's looking for the gaps in your business. These are the questions it asks.

Give me a brain dump of what you do Monday to Friday - what a typical
day looks like for you.

What do you do on repeat - every day, every week, every month? Walk me
through the steps of the ones that take the most time.

Give me a brain dump on what you currently outsource to humans.

Give me a brain dump on the things you absolutely hate doing in your
business.

Give me a brain dump on the things you find easy & the things you find
challenging.

What do you sell - your products and services? Is any of them big
enough that it needs its own focus?

Is there anything that has to stay with you, or with a human, no matter
what?

The question about what you do on repeat is the important one. Those are your SOPs, even if they only live in your head, and they're what become your skills.

Answer the way you'd talk to a friend. Voice notes are perfect for this. Have your calendar and your task list open while you do it, because the small jobs are the ones you'll forget.

Step 4 - Read your org chart and your top five

When you've answered everything, the AI Org Mapper shows you your COO at the top, your departments underneath, the employees each department needs and the skills each employee uses.

For every extra employee, it tells you which of the four tests it passes, so you're not building employees you don't need. It also tells you which jobs aren't worth a skill at all, because sometimes it's not worth it, and other times it is.

Then it gives you the top five that will make the biggest difference in your business, the ones to build this weekend. The first one is always a COO, or a variation like a chief of staff.

Read it through and tell it if anything's in the wrong place. When you're happy, it saves the org chart in your AI brain as AI ORG CHART.md. That's your index, and it's the thing I didn't have when I started.

The catch, and the fix

The org chart is only as good as the brain dump. If you rush the questions, you'll map the jobs you think about and miss the ones that take your time without you noticing, and your COO will build the wrong skills first.

The fix is to keep adding to it. Every time you catch yourself doing something that isn't on the chart, open a chat, run /org-mapper again and tell it what you found. It reads the org chart it already saved and updates it. Do the same whenever your business changes, like a new offer or a new type of client, and before you build anything new. You can also ask it straight out whether you need another employee, and it runs the four tests and tells you.

What this looks like as a business

Every business owner you support is looking at the same lists of agents online and wondering which ones to build. Sitting down with them to map their AI org chart, with the departments, the employees, the skills and the guardrails, is the part they can't do for themselves. You're paid for knowing what to build and in what order.

Start with one

Your org chart tells you what to build, and your COO drafts it with you. Now that you know what you need to make, make your first employee, and it's more than likely going to be a COO. Its primary job is to be your right hand and leverage the skills on your org chart. Go to The First AI Employee I'd Build as a Non-Techie Founder: An AI COO, build that one, and hand it your AI ORG CHART.md. Then set up your workflow manager, so you can hand work to your COO from your task manager.

If you haven't got your AI brain set up, that's the foundation for all of this: start here.

this is the free stuff — imagine what we'd build together

Ready to make your business AI-enabled?

Consulting and one-to-one coaching for founders and freelancers who want AI actually working inside their business — not just talked about.