Setting Up an AI Agent Practice Environment (Part 0: Before You Start)

Setting Up an AI Agent Practice Environment (Part 0: Before You Start)
Summary — The hands-on tutorial in this series is written so you can follow along with any AI coding agent, but it’s much easier to follow when everyone starts with the same tool and sees the same screen. This article covers Anthropic’s Claude Code specifically, from installation through confirming it “actually runs.” Writing code starts in the next article, the hands-on tutorial.

So far this series has covered four concept articles (Part 1, Part 2, Part 3, Part 4), and the hands-on tutorial explains that “any AI coding agent works.” That’s true, but when you’re just starting out, picking one tool and following it exactly gets you moving much faster with far less confusion. This article uses Claude Code.

Before You Start — Account and Cost

First, an honest note. Claude Code is not included in the free Claude.ai plan. You’ll need one of the following.

  • A Claude Pro subscription ($20/month) or higher (Pro, Max, Team, or Enterprise)
  • Or an Anthropic Console account (billed by API usage)

If you just want to spend a few hours on this tutorial, Pro is plenty. If you already have Claude Pro, you can start right away with no extra charge.

Need an alternative? The concepts in this series — agents, context, harness, agentic engineering — apply just as well to tools other than Claude Code (Codex, Cursor, and others). If paying for a specific tool isn’t something you want to do, feel free to work through the tutorial with whatever AI coding tool you’re already using — just skip the installation steps below and go straight to the hands-on tutorial.

What You’ll Need

  • A computer where you can open a terminal (a screen where you can type commands) — macOS, Windows, or Linux all work
  • A Claude account, as described above
  • An internet connection

You don’t need to pre-install anything like Node.js. The current native installer works on its own, with no extra dependencies.

Step 1. Open a Terminal

macOS: Press Cmd + Space, type “Terminal,” and launch it.

Windows: Search “PowerShell” from the Start menu and launch it. (It must be PowerShell — the install command below won’t work in Command Prompt/CMD. If you see PS at the start of the prompt, PowerShell opened correctly.)

[Screenshot placeholder: terminal/PowerShell open]

Step 2. Install Claude Code

Paste the command that matches your operating system into the terminal and run it.

macOS / Linux

curl -fsSL claude.ai/install.sh | bash

Windows (PowerShell)

irm https://claude.ai/install.ps1 | iex

You’ll see a few lines of log output while it installs. If no confirmation dialog pops up, that’s normal.

[Screenshot placeholder: install command running/complete]

Step 3. Confirm the Install

Once installation finishes, run this to confirm it installed correctly.

claude --version

Success means it prints a version number. If you get a “command not found” error, close the terminal completely and reopen it, then try again (right after installing, a terminal window sometimes hasn’t picked up the new path yet).

Step 4. Log In

From any folder, run this to launch Claude Code.

claude

The first time you run it, it will prompt you to log in. Follow the prompt — your browser opens, and you log in with your Claude account. Once login completes, you’ll return to the terminal and see a message like “logged in.”

[Screenshot placeholder: browser login screen → terminal after login completes]

After that, running claude alone launches it directly — no login required.

Step 5. Confirm Your First Run (once this works, you’re ready)

Create a folder for the tutorial, move into it, and run claude again.

macOS / Linux

mkdir budget-practice
cd budget-practice
claude

Windows (PowerShell)

mkdir budget-practice
cd budget-practice
claude

You’re ready once you see Claude Code’s welcome screen and can type into the prompt. To do a quick sanity check that it actually works, try typing this.

Create a file called hello.txt and write "setup complete" in it

If the file actually gets created, you’re all set. This budget-practice folder is the same one you’ll keep using in the hands-on tutorial.

[Screenshot placeholder: after hello.txt is created]

Common Problems

Symptom Cause / Fix
Permission denied when running the install command Don’t try to install with admin rights (don’t use sudo). Instead, open a fresh terminal and try again, or check the official docs’ permission-issue guidance.
claude command not found Close the terminal completely and reopen it. If that doesn’t fix it, the install may have failed partway — rerun Step 2.
irm not recognized on Windows You’re running it in CMD. Open a fresh PowerShell window and try again.
Login browser window doesn’t open Copy the link shown in the terminal and paste it directly into your browser’s address bar.

Next Step

Once your environment is ready, move on to the hands-on tutorial — building a budget-summary app and start right from the budget-practice folder you just created.

AI Agent Series — Full Table of Contents

Beyond the series — the two articles below aren’t part of this 5-part series, but pair well with it.

질문이나 지적할 부분이 있으면 문의로 알려주세요.

Questions or corrections? Let us know via Contact.

AI

AI map Ontology

기업 IT·데이터 조직에서 20년 넘게 실무를 해온 사람이 씁니다. 모든 사례는 익명화·일반화합니다. 소개 보기 →

AI

AI map Ontology

Written by someone with 20+ years in enterprise IT and data. All cases are anonymized and generalized. About us →

다음으로 읽어볼 글

개념을 이해했다면, 실제 설계와 활용 방법을 이어서 살펴보세요.

온톨로지 Foundry AIP 기업 AI 전략

Keep reading

Once you understand the concept, continue on to real design and usage patterns.

Ontology Foundry AIP Enterprise AI Strategy