# AI 101: set up your AI stack

**The CV Method · AI 101 · Lesson 1 · Working with AI · Aug 2026**
*by Corina · @the_cv_method · hello.thecvmethod.com · current as of Aug 2026 — the AI space moves fast, so double-check tool names*

The complete-beginner's guide to getting AI set up so it actually helps — a tidy home for your files, a short note that tells the AI how you work, and a few simple rules. **Plain English, no coding, about an afternoon.**

---

## What you're building

A small, calm setup you actually understand: **one AI**, **one home folder**, and a short "read-me-first" note (it's called a **CLAUDE.md**) that tells the AI who you are and how you like things — so it stops guessing and starts helping. That's the whole "stack." No jargon you can't explain, nothing you can't undo. It rests on one rule.

## Rule zero — one tool, one folder. Add the rest later.

You don't need forty apps. You need one AI you trust and one place it can read. Everything else you add only when a real task asks for it — never before.

## Step 1 · Pick one AI (and what "stack" means)

Your "AI **stack**" is just the few tools you use together — the AI itself, plus wherever your files live. That's the whole word. Pick **one** capable AI and stop there for now.

| Piece | What it is (plain) | I use | Or |
|---|---|---|---|
| **The brain** | The AI that reads and writes for you. | Claude | Any strong AI assistant |
| **Where files live** | A place the AI can read your stuff. | A notes vault (plain files) | A folder on your computer · Notion |
| **Helpers (later)** | Connectors, schedules, saved "skills" — only when needed. | added one at a time | — |

*Not sure? Start with the AI you already have open. The tool is the easy part — the setup below is what makes it useful.*

## Step 2 · Make one home folder

Make **one** folder that all your AI work lives in. Inside it, a few boringly-named boxes. Folders are just labeled boxes — keep the names obvious so you (and the AI) always know where things go.

```
my-ai/
  context/    ← who you are + how you work (the AI reads this)
  projects/   ← one folder per thing you're working on
  inbox/      ← drop files here to work on
  done/       ← finished stuff, out of the way
```

*Boring and obvious beats clever. You should be able to find anything in about two clicks.*

## Step 3 · Write a CLAUDE.md (your "read-me-first" note)

A **CLAUDE.md** (say "Claude-dot-M-D") is just a plain text file where you tell the AI, **one time**, who you are, how you like things, and your rules. Claude Code reads it automatically; in the Claude app or ChatGPT, paste it in or put it in a Project's instructions — so you stop repeating yourself. *(It's named for Claude, but the idea — an "about me + house rules" note — works with any AI; some just call it a "system" or "context" doc.)*

```
# About me
I run a small business. I'm not technical — keep it simple.
# How I like things
Plain English. Short. Ask before you assume.
# My rules
- Draft, don't send. I approve everything.
- Never share private info or make up numbers.
# Where things live
Background in /context · active work in /projects.
```

*Keep it to about half a page. You'll add to it over time — every time you catch yourself repeating something, put it in here.*

## Step 4 · Give it context, not your whole life

AI is only as good as what it can see. For any task, point it at the **specific** folder or file — not everything you own. Think of it this way: your **context/** folder and CLAUDE.md are the always-on background; the **project folder** is the "right now."

*Too much info makes AI wander; the right info makes it sharp. When in doubt, give it less — then add more if it needs it.*

## Step 5 · Set your ground rules

Write your non-negotiables into the CLAUDE.md so they're **always on**: it **drafts, you approve**; nothing sends, posts, or pays on its own; keep private info out; when unsure, ask. Boring rules are exactly what make AI safe to hand real work to.

*"It proposes, you approve" is the single rule that prevents almost every AI horror story you've heard.*

## Step 6 · Grow it slowly

Resist installing forty things. When a **real task** needs something new — a connector to your email, a scheduled run, a saved "skill" — add **that one thing**. A small setup you understand always beats a big one you don't.

*The goal isn't the fanciest stack. It's a calm one you actually use.*

## Keep it honest & safe

- It proposes, you approve.
- One tool to start.
- Plain files you own.
- Keep private info out.
- No invented numbers.
- Date it — AI moves fast.

## Steal this prompt

**The setup helper — paste and fill in the blank**

```
Here's who I am and how I work: [a few sentences].
Help me set up a simple AI workspace. Suggest a plain folder structure,
and draft a CLAUDE.md I can edit — an "about me", how I like things, and
my rules (draft-don't-send; no private info; no made-up numbers).
Keep it plain English, no jargon, and ask me anything you need first.
```

## The starter checklist

1. Pick **ONE** AI to start.
2. Make a home folder: **context / projects / inbox / done**.
3. Write a short **CLAUDE.md** (about me · how I like things · my rules · where things live).
4. Point the AI at the **right folder** per task — not everything.
5. Put **"draft, don't send"** + your rules in the CLAUDE.md.
6. Add a new tool **only** when a real task needs it.
7. Keep **private info** and secrets out.
8. Revisit the note as you learn what you repeat.

---

**Want your whole AI foundation set up for you?** I'll build the folders, the rules, and the workflows for your business — so you skip the overwhelm and start from "it just works." Start at **hello.thecvmethod.com**.

*The CV Method · AI 101 · Lesson 1 — set up your AI stack*

---

*Part of the **AI 101** series: Lesson 1 — Set up your AI stack · Lesson 2 — Your AI roadmap · Lesson 3 — How to prompt · Lesson 4 — Give it your context. More lessons at hello.thecvmethod.com/playbooks.*