Skip to content
View on GitHub

Get Claude Code

Claude Code

Claude Code is Anthropic's agentic coding tool. It runs in your terminal, reads and edits files in your project, runs commands, and iterates: you describe what you want in plain language and it does the work, asking before anything risky.

Install

curl -fsSL https://claude.ai/install.sh | bash

Then, from inside a project folder:

claude

Other ways in, if you'd rather stay in an editor:

Full docs: code.claude.com/docs.

Desktop app

The Claude Desktop app (macOS, Windows, Linux beta) wraps the same agent in a GUI, under its Code tab. Install, sign in, click Code, and open a session against a project folder. Compared to the terminal:

  • Panes, not just text: chat, diff viewer, integrated terminal, and file editor arranged side by side, plus a browser pane where Claude previews and clicks through your running app to verify its own changes.
  • Permission modes in a dropdown instead of flags: Manual (asks before every edit), Accept edits, Plan (read-only, proposes first), Auto, and Bypass (same underlying settings as the CLI's permissions.defaultMode).
  • Parallel sessions: each new session gets its own git worktree, so you can run several tasks on the same repo at once without them colliding.
  • Diff review: comment on specific lines before Claude proceeds, or click Review code to have Claude self-review its own diff for bugs before you commit.

Full reference: code.claude.com/docs/en/desktop.

Using its agent on a project

  1. cd into your project folder and run claude.
  2. Paste the project brief from Foundations, Clinical Application, or Basic Science and ask it to scaffold the app/notebook.
  3. Let it propose a plan before it touches files. Claude Code asks for confirmation on anything destructive (installs, deletes, pushes).
  4. Iterate: point it at failing tests or unexpected output and ask it to fix the root cause, not just silence the symptom.

← Back to home