The 5 Minute Setup That Turned My Terminal Into an AI Engineer
I opened my laptop late at night telling myself I would just check one thing before going to sleep. That is the classic lie developers tell…
The 5 Minute Setup That Turned My Terminal Into an AI Engineer
I opened my laptop late at night telling myself I would just check one thing before going to sleep. That is the classic lie developers tell themselves. Five minutes later you are supposed to close the laptop, but somehow two hours pass and you are still there, typing prompts into your terminal while Claude reads your repository and suggests fixes faster than you can scroll through the files.
If you have not tried Claude Code inside your terminal, it is one of the easiest upgrades you can make to your workflow. Instead of constantly switching between your editor, documentation, and search results, you can just ask questions directly in the terminal while Claude reads your project files.
This guide shows how to set it up in about five minutes so you can start using it inside your own projects.
Step 1: Check Your Claude Plan
Before installing anything, make sure your Claude account supports Claude Code. The feature is currently available only on paid plans, so if you are on the free plan the installation will work but authentication will fail later. This is the most common point where people get stuck.
Go to claude.ai, log in, and check your account plan. You should see something like Claude Pro, Max, Teams, or Enterprise listed in your account settings. If the plan shows as Free, upgrade first and then continue with the setup.
Once you confirm the account has a supported plan, you are ready to install Claude Code.
Step 2: Install Claude Code
Open your terminal and run the installation command. This downloads and installs Claude Code automatically.
curl -fsSL https://claude.ai/install.sh | bash
Your terminal will start printing installation logs as the script runs.
Downloading Claude Code...
Installing dependencies...
Configuring environment...
Claude Code installed successfully
If the installation finishes without errors, the tool is now installed on your system.
Step 3: Verify the Installation
Next, confirm that the command works properly. This step ensures your terminal recognizes the Claude executable.
claude --version
If everything is installed correctly, you should see a version number printed.
claude 1.x.x
If you see command not found, simply close the terminal and open a new one so the environment reloads. After restarting the terminal, run the command again and the version number should appear.
Step 4: Log In
Now you need to authenticate your Claude account so the terminal tool can connect to your subscription.
claude
When you run this command for the first time, Claude will ask you to log in through your browser.
Opening browser for authentication...
Please log in to your Claude account
Authentication successful
Connected to Claude
A browser window should open automatically. Log in using the same account that has your paid plan. Once the login is completed, the terminal session will confirm the connection.
Step 5: Open Your Project Folder
Claude works best when it understands your repository. That means you should run it inside the folder containing your project instead of your home directory.
Navigate to your project folder first.
cd my-project
Then start Claude again.
claude
Claude will scan the project and load the context.
Scanning project files...
Indexing repository...
Project loaded successfully
You can now ask Claude about your code
After this step, Claude understands your codebase and is ready to answer questions about it.
Step 6: Add a CLAUDE.md File
To make Claude even more useful, create a small file called CLAUDE.md in your project root. This file tells Claude how your project works and what commands or rules it should follow.
Example:
## Commands
npm run dev
npm test
## Notes
TypeScript strict mode enabled
Do not modify the legacy auth folder
Whenever Claude runs inside your repository, it reads this file automatically. That means you do not have to explain your project structure every time you start a session.
Step 7: Ask Claude Something Real
At this point everything is ready and you can start interacting with Claude directly in the terminal. Instead of searching documentation or digging through stack traces alone, you can ask questions about your actual code.
For example:
Why is this function returning undefined?
Claude reads the relevant files and responds with suggestions based on your project.
Analyzing function...
Possible issue found in userService.ts line 42
Variable may be undefined before return
Suggested fix available
This is usually the moment when it clicks. Instead of generic answers, Claude is responding with context from your repository.
Final Thoughts
The setup takes only a few minutes, but the workflow change is noticeable almost immediately. You stay inside your terminal, ask questions about your code, and get responses that understand your project instead of generic explanations.
Once you start using it regularly, it becomes hard to go back to the old workflow of constantly switching between tools. The terminal stops being just a place to run commands and starts feeling more like a conversation with your codebase.
If you followed the steps above, Claude Code should now be running in your terminal and ready to help you build faster.
메타데이터
- post_id
- aedd334fffec
- slug
- the-5-minute-setup-that-turned-my-terminal-into-an-ai-engineer-aedd334fffec
- url
- https://medium.com/@imvk45/the-5-minute-setup-that-turned-my-terminal-into-an-ai-engineer-aedd334fffec
- canonical_url
- https://medium.com/@imvk45/the-5-minute-setup-that-turned-my-terminal-into-an-ai-engineer-aedd334fffec
- author_url
- https://medium.com/@imvk45
- status
- ok
- fetched_at
- 2026-06-14 11:28:49