Claude Code — how to program without writing code
A practical step-by-step course for non-technical people: from installing Claude Code in the terminal, through connecting your account and API, to building and launching your first app without writing code yourself.
This course walks you through the real workflow of working in Claude Code: installing the tool, logging in, basic terminal commands, configuring permissions and settings, working with slash commands, connecting an API key, and creating and launching your first simple app. Instead of theory, you get a series of concrete actions in the terminal and within the project structure: you create a project folder, launch Claude Code with the `claude` command, use `/init`, `/config`, `/permissions`, `/model`, `/cost`, and `/doctor`, save rules in `CLAUDE.md`, protect files like `.env` in `.claude/settings.json`, and then build a simple app with AI’s help. The course reflects Claude Code’s current workflow: installation via npm, running it in the terminal, interactive login through `/login`, and using `ANTHROPIC_API_KEY` for API- and SDK-based work. By the end of the course, you’ll be able to assign Claude Code the next steps in building a product, run the app locally, and improve it without manual coding.
What you will learn
- You will install Claude Code and run it in the terminal in your own project folder.
- You will perform the basic terminal commands needed to work without a programming background.
- You will configure login, model, permissions, and project settings in Claude Code.
- You will connect Claude Code to your Anthropic account and learn the difference between interactive login and using an API key.
- You will prepare a secure working environment with `CLAUDE.md`, `.claude/settings.json`, and `.env` files.
- You will give Claude Code a series of commands that lead to creating your first working app.
- You will run the app locally, check for errors, and ask Claude Code to fix them step by step.
- You will build your own simple workflow for creating additional mini-products without writing code yourself.
Prerequisites
A computer with macOS, Linux, or Windows with WSL; Node.js 18+ installed; access to a terminal; an Anthropic account; optionally an Anthropic API key for the section on working via API; basic ability to copy commands into the terminal.
Course syllabus
- Claude Code is not a chatbot: what it actually does in the terminal and on project files
- Step-by-step installation: Node.js 18+, `npm install -g @anthropic-ai/claude-code`, and first launch of `claude`
- Logging into an Anthropic account and starting a session in the project directory without losing context
- First no-code task: ask Claude Code to create a simple app sketch from ready-made files
- Comparison of full prompts: an overly vague prompt vs. a prompt Claude can actually use to do the work
- Task prompt template for no-code work: goal, constraints, input files, output, and acceptance criteria
- Brief workshop: how to describe a CRUD app, landing page, or automation script without knowing the syntax
- One big prompt or a series of small steps? A decision using the example of building a mini app
- Comparison of artifacts: weak feature description vs good feature description with an acceptable final result
- How to ask for a plan before changes: first repository analysis, then implementation
- How to make Claude Code understand a repository first: analyzing files, dependencies, and entry points
- Adding a new feature without coding: from business request to ready changes in the project
- Fixing a bug based on a symptom: how to describe the problem so Claude finds the failure point on its own
- Refactoring Without Touching Syntax: Asking to Simplify Code and Preserve Behavior
- Reading diffs and assessing risk: what to check before you accept Claude Code edits
- `/init` and `CLAUDE.md`: saving project rules, commands, and working style so you don’t have to repeat instructions
- `/memory` and context updates: how to add lasting decisions after each iteration
- `.claude/settings.json` in practice: available directories, permission modes, and blocking sensitive files
- AI work safety in the terminal: what not to share and how to limit access to secrets
- Hooks Without Magic: When a Simple Hook Helps, and When a Beginner Should Stick to Manual Control
- `claude -p` instead of an interactive session: when to run a task with one command
- Output `text`, `json`, and `stream-json`: how to receive results for further use in tools and automations
- Mini-workshop: ask Claude Code to analyze a project and generate a checklist of changes to implement
- GitHub Actions with `@claude`: how to run a simple automation in a repo without building your own agent
- Final quiz: choosing the best workflow for creating, improving, and automating no-code applications
FAQ
The course guides you through the real, practical workflow of working with Claude Code: from installation and login, through basic terminal commands, to configuring permissions, settings, and models. Then you learn how to use commands such as /init, /config, /permissions, /model, and /cost, connect an API key, and create and run your first simple app. It’s learning by doing, not by theory.
Yes. It’s a good choice for beginners who want to enter the world of app creation with AI without learning traditional coding from scratch. The course shows concrete steps in the terminal and project structure, so you can understand the process faster and start working independently.
No. The course introduces the basic terminal commands needed to work with Claude Code, so you only learn what will actually be useful in practice. Instead of overwhelming theory, you get a simple, organized process.
Because AI coding tools are quickly becoming a standard part of work and learning. According to the Stack Overflow Developer Survey 2025, OpenAI tools were used in development work by 82% of respondents, and 36% of those surveyed were learning to use AI tools to advance their careers. At the same time, many users still prefer simpler forms of support instead of advanced agents, which makes a practical course based on a concrete workflow especially valuable today.
Above all, it’s concrete. Instead of vague talk about artificial intelligence, you get a sequence of actions that you perform step by step: you create a project folder, launch Claude Code with the `claude` command, configure the environment, and immediately move on to your first working result. This course is for people who want to get up to speed faster with a new way of working with AI.
After completing the course, you will understand how to launch and configure Claude Code, how to navigate the basic terminal workflow, and how to use the tool to create and run your first simple app. You’ll gain a practical foundation for further developing projects with the help of AI.
- 6 hours
- Beginner
- Certificate on completion
- Access immediately after purchase