Claude Code
Definition
Claude Code is Anthropic’s agentic coding tool: software that works with a real code project, reads the files already there, and can edit code, run commands, and handle git-style workflows from plain-language instructions. The terminal is the classic version, but Anthropic also offers it in IDEs, desktop, web, Slack, mobile handoff, and automation surfaces. The important distinction is this: ordinary Claude chat can talk about code; Claude Code is built to do development steps in the project folder, with permission modes that control when it may change files or run tools.
Why It Matters
Once you know Claude Code is a hands-on coding agent in your project — not “Claude the chat window with a new nickname” — tool lists and “we use Claude Code” disclosures stop sounding like jargon. You can also hear the real tradeoff: it is powerful because it can touch files and run commands, which is exactly why approval modes and careful use on important folders matter.
What does that look like in practice?
Your cousin left town and handed you a half-finished bake-sale signup page for the school PTA: a messy folder of HTML, a half-working form, and a README that says “TODO: make the email actually send.” You open a terminal in that folder, start Claude Code, and say: “Explain what this project does, then fix the form so it validates email addresses and shows a clear thank-you message — show me the plan before you edit.” It maps the files, proposes the change, waits for your go-ahead if your permission mode requires it, edits the right spots, and can commit when you ask. Natural language in, real project work out. Slightly unnerving, yes. Also rather useful.
What people actually mean when they say this
When someone says “I did it in Claude Code,” they usually mean they ran Anthropic’s coding agent against a real codebase — in the terminal, an IDE, web, desktop, Slack, or another connected surface. They do not mean they only chatted on claude.ai, pasted code into a generic chatbot, or made a Claude Artifact in a side panel.
Not to be confused with
Last reviewed: August 2026
Related reading
In the Privacy Decoder
- ClaudeAI assistant from Anthropic. Everyday conversation, writing, analysis, code, and long documents, plus Artifacts: side-panel documents and small interactive tools you can edit, download, and optionally publish.Medium
- Hermes AgentAn open-source personal AI agent that runs on your own computer, remembers your projects over time, and reaches you through messaging apps. A close cousin of OpenClaw.Low
- OpenClawAn open-source personal AI agent that runs on your own computer and connects to your apps. You message it through WhatsApp, Telegram, or Discord, and it carries out tasks for you.Low
Knowing the word is just the start.
We help non-tech people go from looking things up to actually using AI confidently. Tutorials, prompts, plain English. All of it, for less than a cup of coffee a month.
Show Me How This Works →Start free. Cancel anytime. No judgment, ever.