# Building an agentic brain, from zero to measured | Team0

> A build-along series by Yogev Ben-Tov. Practical exercises and a companion skill to help you build and measure your own agent memory, from MD files to databases.

Agent-readable version of https://team0.ai/learn. The page for people is at that address; this file carries the same content without layout.

---
# Building an agentic brain.

From zero to a brain your agents read from, measured at every step. Build in small steps, make your own design choices, and score your agents after each change. MD files, databases and knowledge graphs are all welcome.

## The series.

1. [Part 0

   ### Measure before you build

   Real questions, expected truth, a baseline and a held-out set, for any memory format](https://team0.ai/learn/measure-before-you-build)
2. [Part 1

   ### Instruction files, and a first awareness layer

   A small source-backed prototype and clear boundaries around the information it keeps](https://team0.ai/learn/instruction-files-and-a-first-awareness-layer)
3. [Part 2

   ### Facts and entities

   A starter representation of people, work and sourced statements](https://team0.ai/learn/facts-and-entities)
4. [Part 3

   ### The write path

   Rules you choose for uncertainty, changed values and corrections, tested on real cases](https://team0.ai/learn/the-write-path)
5. [Part 4

   ### Keeping junk out

   A quality check for useful information and samples of what your system misses](https://team0.ai/learn/keeping-junk-out)
6. [Part 5

   ### Keeping it true over time

   An exercise in completed promises, disagreements or the consequences of a change](https://team0.ai/learn/keeping-it-true-over-time)
7. [Part 6

   ### The read path

   A useful, inspectable read for your tasks, storage format and budget](https://team0.ai/learn/the-read-path)
8. [Part 7

   ### Making agents use it

   One host connected and tested for reading, saving and failure behavior](https://team0.ai/learn/making-agents-use-it)
9. [Part 8

   ### More agents and more people

   A sharing prototype, an access policy and tests for useful sharing and privacy failures](https://team0.ai/learn/more-agents-and-more-people)
10. [Part 9

    ### Running it

    Health checks, cost and comparable measurements as the system changes](https://team0.ai/learn/running-it)

Each part gives you an exercise, the decisions to make and a way to test the result. You can stop at any point and keep what you built.

## Your agent can walk you through it.

The companion skill follows this series: explain a part, help write your implementation, review your existing setup, or measure the result. Its scoring method works with any memory format, including MD files.

Install the companion skill from its [open-source repository](https://github.com/Team0-ai/agentic-brain-design):

```
npx skills add Team0-ai/agentic-brain-design
```

The skill includes a scoring tool: give it test questions, the expected answers and your agent’s answers to compare results before and after a change. [See how scoring works](https://team0.ai/learn/measure-before-you-build).

## A company brain you can buy.

Team0 is a maintained company product installed in your cloud, with updates and support. Your company controls the infrastructure and provider accounts. [Explore the Team0 product](https://team0.ai/product).