Skip to main content

Getting Started with AI Tools

Getting Claude Code working takes 30 minutes. Getting it working well takes one more hour. This guide covers both.

By the end you'll have an AI that knows your standards, follows your voice, and applies your judgment automatically — without you typing the same instructions every session.

Before You Start​

You need two things:

  1. Claude Code installed — run claude --version in your terminal. If you see a version number, you're set. If not, install it: npm install -g @anthropic-ai/claude-code (requires Node.js).
  2. A project folder — any folder with files you want to work on. A content folder, a code project, even a collection of documents works.

Step 1 — Open Claude Code (5 min)​

Navigate to your project folder in the terminal:

cd /path/to/your/project
claude

You'll see the Claude Code prompt. Type anything — ask it to summarize the files in the folder, or just say "hello." This confirms it's working.

What you're looking at: An AI with access to your files. It can read, edit, create, and run commands. Nothing happens without you asking — but it can do a lot when you do.

Step 2 — Write Your First CLAUDE.md (20 min)​

CLAUDE.md is a file you create at the root of your project. Claude reads it automatically at the start of every session. It's how you teach the AI how you work.

Create the file:

touch CLAUDE.md

Open it and add this minimum viable template:

# Project Instructions

## What This Is
[One sentence describing your project]

## How to Write
- [Your voice or style — e.g. "Short sentences. No jargon."]
- [Any formatting rules — e.g. "Use headers for sections. No bullet soup."]

## What to Avoid
- [Common mistakes you don't want repeated]
- [Any off-limits actions — e.g. "Never delete files without asking."]

## Context
[Two or three sentences about who this work is for and what success looks like]

Fill it in honestly. This file compounds — the clearer it is, the better every session gets.

For a creative project (content, marketing, media): Focus the "How to Write" section on voice, tone, and what good looks like. "Write like Hemingway. Concrete nouns. Active verbs. No filler." is a valid rule.

For a code project: Focus on patterns, what frameworks you're using, and what standards to enforce. The Config Architecture guide goes deeper.

Step 3 — Add One Rule (5 min)​

Rules are always-loaded constraints — things you never want Claude to forget. They live in .claude/rules/ and are imported by your CLAUDE.md.

Create your first rule:

mkdir -p .claude/rules

Create .claude/rules/voice.md:

# Voice

Write short. Use the active voice. No pleasantries ("Great question!"). 
No hedging ("It seems like..."). Direct answers only.

Then add this line to your CLAUDE.md:

@.claude/rules/voice.md

The @ prefix tells Claude to load that file every session. Now your voice rule is always active — you don't have to repeat it.

Test it: Ask Claude to write a paragraph about your project. If it's still wordy or hedging, tighten the rule. Iteration is fast.

Step 4 — Run Your First Skill (5 min)​

A skill is a reusable workflow — a set of instructions Claude follows when you invoke it with a slash command. Skills are how you turn repeatable work into one-line commands.

Claude Code ships with built-in skills. Try one:

/commit

This skill reads your git changes and writes a commit message following best practices. No instructions needed — the skill contains them.

Other built-in skills to explore:

  • /fix-links — finds and repairs broken internal links
  • /review-pr — reviews a pull request and flags issues

You can also build your own skills. A skill is a markdown file. Any instruction you find yourself typing repeatedly is a candidate. See Skills for how.

Step 5 — The Cascade Is Running (ongoing)​

You now have:

  • A CLAUDE.md that orients every session
  • A rule that enforces your voice
  • Skills you can invoke for repeatable work

That's the beginning of the Cascade:

Mantra (your intention)
  → Rule (always enforced)
    → Hook (runs automatically on every action)
      → System (you never think about it again)

You're at Rule. The next step — hooks — are shell scripts that run when Claude takes certain actions. They're optional but powerful: a hook can automatically check that every file Claude edits passes a quality test, without you asking.

The pattern compounds. The longer you run it, the more it learns your judgment. The more it learns your judgment, the less you explain.

What's Next​

If you want to...Go here
Understand the full config architectureConfig Architecture
Build your own skillsSkills
Use subagents for parallel workAgents
Choose the right AI modelLLM Comparison
See the Cascade concept in fullMantra

Context​

  • Config Architecture — how to structure your config so it works with any AI tool, not just Claude
  • Claude Code Deep Dive — the full setup guide including hooks, memory, and agent teams
  • Mantra — the Cascade that underlies every productive AI setup
  • Skills — modular workflows you can invoke or compose
  • Standards — the AAIF standard that governs how skills and agent config are structured

Questions​

What would have to be true about your CLAUDE.md for Claude to be genuinely useful after one session, without any additional instruction?

  • Which of your repeatable workflows would save the most time if you could invoke them with a single slash command?
  • If the Cascade fully runs — Mantra → Rule → Hook → System — what does your work look like differently in six months?
  • Where does your judgment matter most in your current work, and is that where you want AI to learn first?