Loading interactive tool…

This page in plain text

Claude Code: what it is good at and how to start

Anthropic’s coding agent. You describe what you want in plain words and it does the work in your project: it reads the files, runs commands, makes the change and checks it, showing you each step. This site was built with it.

Best for: building and changing real software. See Claude Code.

What it is good at

  • Working in a real project. It reads the code that is already there before it changes anything, so the change fits.
  • Checking its own work. It can run the tests, open the page and look, and tell you what it could not verify.
  • Taking on know-how. Skills and MCP servers give it your team’s methods and access to your tools.

Try this first: a first build

Build [thing] for [who uses it].

It is working when [the check you would do by hand].

Before you write any code, read the project and tell me your plan in five lines.

When you finish, show me it running, list what you changed, and tell me anything you could not verify.

What to use Claude Code for

All twelve jobs are on our solutions page.

How Claude Code got here

  • 25 November 2024: The Model Context Protocol. An open standard for connecting assistants to the places your data lives: files, business tools, databases. The plumbing that lets an assistant work with your own material.
  • 24 February 2025: Claude 3.7 Sonnet and Claude Code. Claude Code arrived as a preview: an agent that works in your project from the command line, reading files, running commands and making changes.
  • 22 May 2025: Claude 4. Models built to keep working on a long task for hours, and Claude Code became generally available. Delegating a whole job, not a single step, came into reach.
  • 16 October 2025: Agent Skills. Folders of instructions, scripts and resources that Claude loads only when a task calls for them. Know-how became something you write once and reuse.

More on the AI timeline.

Learn Claude Code properly

  • Claude Code in Action, by Anthropic. How to work with a coding agent in a real project. This site was built with Claude Code, so we would start here.
  • Anthropic’s courses on GitHub, by Anthropic. Anthropic’s own notebooks on the API, prompting, evaluating prompts and tool use. Run them and change them as you go.
  • Official AI skills. Folders of know-how it loads when a task calls for them.
  • Official MCP servers. Connections to your files, browser and business tools.

What to watch

  • It runs commands on your machine. Read what it proposes, work on a branch, and keep anything it cannot undo behind your own yes.
  • “Done” is a claim until you have seen it running. Ask it to show you.

Claude Code or Codex?

We use both, and Claude Code does most of the building. Codex is our second opinion when Claude Code gets stuck, and what we add when we run through our Claude plan’s usage. If you are starting out with a coding agent, start with Claude Code. The full comparison. Read about Codex.

Other tools we use: ChatGPT, Claude, Codex, The Claude API, The OpenAI API, AI Engine, ElevenLabs, Gemini, Perplexity, NotebookLM, Midjourney. Or see all tools.

Ask Another AI Guy

I'm just Another AI Guy.

You're chatting with our fast model. Create a free account for the more thorough one, and more messages.