Claude Code
In short
Claude Code is Anthropic's AI coding agent: it reads your codebase, edits files, runs commands and works with git in a terminal, IDE, desktop app or browser.
What is Claude Code?
Claude Code is a coding agent made by Anthropic and powered by its Claude models. Instead of suggesting one line at a time, it takes a task described in plain language, such as 'add password reset to the login page', explores the project to understand it, plans the change, edits the files involved and runs commands such as tests to check its work. It started as a command-line tool and now also runs as extensions for VS Code and JetBrains IDEs, in a desktop app and in the browser, where tasks run in the cloud.
A project gives it standing instructions in a CLAUDE.md file, which it reads at the start of every session; it can also read an AGENTS.md file written for other agents. It connects to outside tools and data through the Model Context Protocol, and teams extend it with skills (packaged workflows), hooks (shell commands that run before or after its actions) and subagents that handle parts of a task in parallel. By default it asks before editing files or running commands.
Working with it is closer to handing a task to a capable colleague than to using autocomplete: you describe the goal, it does the legwork across many files, and you review the result much as you would review a teammate's pull request.
Claude Code is often confused with Claude, the chat assistant. Both use the same models, but Claude Code is built for software work: it runs where your code is, can change files and run commands, and is scriptable, so it can also run without a person in CI, for example to review pull requests. Most ways of using it need a paid Claude plan or an Anthropic Console account.
Key takeaways
- Claude Code is Anthropic's AI coding agent, powered by Claude models.
- It reads code, edits files, runs commands and creates commits and pull requests.
- It runs in a terminal, in VS Code and JetBrains IDEs, in a desktop app and in a browser.
- A
CLAUDE.mdfile in the project gives it standing instructions. - It connects to other tools through MCP and can run unattended in CI.
Example
cd my-project
claude # start an interactive session in this project
# Or run a single task and print the result, e.g. in CI
claude -p "run the tests, fix any failures and summarize what you changed"Readers ask
Is Claude Code free?
No. It needs a paid Claude plan or an Anthropic Console account billed by usage. The terminal tool and the IDE extensions can also reach Claude models through third-party cloud providers.
Which programming languages does Claude Code work with?
Any language in your repository. It reads the code as text and runs your project's own tools, such as its test runner, linter and build, so it works with whatever stack the project uses.
Is it safe to let Claude Code run commands?
By default it asks before editing files or running commands, and you can allow, deny or sandbox specific actions. Give it only the access it needs and review its changes like any other code before merging.
See also
- Claude Code vs CodexClaude Code and OpenAI Codex are both AI coding agents; they differ mainly in the models behind them, where they run and how you set them up and pay for them.
- AI Coding AgentAI & Machine Learning, p. 5An AI coding agent is an AI system that takes a programming task, explores the code, edits files and runs commands in a loop until the task is done.
- OpenAI CodexAI & Machine Learning, p. 45OpenAI Codex is OpenAI's AI coding agent: it writes, edits and reviews code from a command line, an IDE extension or ChatGPT, on your computer or in the cloud.
- CursorAI & Machine Learning, p. 20Cursor is an AI code editor built on VS Code, with AI autocomplete, a chat and coding agents that edit across files and can use models from several AI labs.
- GitHub CopilotAI & Machine Learning, p. 29GitHub Copilot is GitHub's AI coding assistant: it suggests code as you type, answers questions in chat and, as an agent, turns issues into pull requests.
- Model Context ProtocolAI & Machine Learning, p. 40The Model Context Protocol is an open standard that defines how AI applications connect to external tools, data sources, and prompts through a shared interface.
- Vibe CodingAI & Machine Learning, p. 68Vibe coding is building software by describing what you want to an AI and accepting the code it writes, mostly judging the result by whether it seems to work.
Free · once a week
Software and AI, explained once a week
The week's software and AI news with every term explained, plus a term to learn and a quick quiz. A two-minute read.
See this week's issueAbout this page
Spotted a mistake or something missing on this page?Suggest an edit
How pages are made