Start here
Tier 0

Start here

Never used Claude Code? Install it, sign in, run a first session, make a first edit, and learn to read what it shows you.

8 articles31 min total

Start reading →

  1. What Claude Code is, and what a harness isA model talks. A harness gives it a loop, tools, files and a way to ask you first. Claude Code is the harness these docs teach with.3 min read
  2. Install Claude CodeInstall the CLI on macOS, Linux or Windows, confirm it works with claude --version, and keep it updated.3 min read
  3. Sign in: subscription or API keyPro, Max, Team or a Console API key: which to pick, how billing differs, and the stray-key mistake that quietly costs money.4 min read
  4. Your first sessionOpen Claude Code in a folder, ask for something small, watch it read and act, and exit cleanly.4 min read
  5. The working directory and your first editHow Claude Code reads and writes files in your project, and how to review a change before you accept it.4 min read
  6. Permissions and safety basicsPermission prompts, allow rules, permission modes, what skipping permissions really means, and why destructive actions get a human yes.5 min read
  7. Reading the output: plans, tool calls, resultsHow to read what a session shows you, spot when the agent is stuck or guessing, and step in before it wastes an hour.4 min read
  8. Where to go next: learning pathsPick a path by goal: first week, one agent to many, how a production AI org runs, or a PM's path. Ordered reading lists.4 min read