🤖 HarDojo
Log in Sign up

Claude Code

Anthropic's terminal agent — install, basic commands, and workflow.

🧪 Try it in the Playground 🎯 Practice this path

Claude Code is Anthropic's agentic coding tool that runs in your terminal. It reads your repo, edits files, runs commands, and talks to you in plain English.

Install

npm install -g @anthropic-ai/claude-code
claude

On first run it asks you to sign in to your Anthropic account (or set ANTHROPIC_API_KEY).

Core workflow

claude                      # start in the current directory
claude "add dark mode"      # one-shot task
claude -c                   # continue the last session
claude -p "explain app.js"  # print mode — no interactive UI

Inside the REPL, type requests naturally:

> There's a bug where login fails when the email has uppercase letters. Find and fix it.

Claude will search, propose a plan, edit files and run tests — asking permission for risky actions depending on your settings.

Useful flags and config

  • claude --continue — resume previous conversation.
  • claude /init — generate a CLAUDE.md that documents your project.
  • CLAUDE.md — project memory. Claude reads it automatically; put build commands and conventions there.
  • /config — switch permissions mode (ask / auto-approve / plan-only).

Permissions matter

Claude Code has a permission system for file edits and shell commands. Keep it on ask until you trust it, then allow specific commands. Never run it with auto-approve inside a directory containing secrets.

Practice: run claude /init in a real repo and read the generated CLAUDE.md — it's a great template for all agents.
";