Claude Academy
Sign in

Track 4 — Claude Code

Claude Code is Claude as an agent in your terminal and editor — it reads and writes files, runs commands, and works across your whole repo. This track takes you from first install to genuinely productive: configuring projects with CLAUDE.md, custom slash commands, hooks, plan mode, and the agentic workflows that make Claude Code a daily driver. It closes with Cowork, where Claude works alongside you on broader tasks.

Audience: developers. · Est. time: ~8–10 hours over a week. · Prereq: track-3-building-with-the-api and the dev toolchain incl. Claude Code installed (setup-checklist, section c). · Outcome: you're productive in Claude Code on real work.

Sequence

1. Claude Code 101 · ⏱ ~2.5 h + project

  • Course: claude-code-101 — install and sign in, the interactive session, asking Claude to read/edit/run, the permission model, plan mode, and the role of a CLAUDE.md project memory file.
  • Project: p10-claude-code-onboarding — point Claude Code at a real repo, write a useful CLAUDE.md, and complete a real change (a fix or small feature) start to finish.

2. Claude Code in Action · ⏱ ~3–4 h + project

  • Course: claude-code-in-action — going deeper: custom slash commands, hooks (e.g., auto-run a linter on PostToolUse), settings and permissions, subagents inside Claude Code, multi-file refactors, test-driven workflows, and using Claude Code in CI.
  • Project: p11-custom-command-and-hook — add a custom slash command and a hook that automatically runs your linter or tests after edits.

3. Cowork — Claude alongside you · ⏱ ~2 h + project

  • Course: introduction-to-claude-cowork — Cowork as a shared workspace where Claude collaborates on a task with you, beyond single-message chat. Where it fits relative to chat and to Claude Code.
  • Project: p04-cowork-workspace — set up a Cowork workspace and drive one real task through it end to end.

You're ready to move on when…

  • You can start Claude Code in a repo, use plan mode, and approve/deny actions deliberately.
  • You've written a CLAUDE.md that measurably improves Claude's behavior in your project.
  • You can author a custom slash command and a hook, and explain when each fires.
  • You can describe how Cowork, claude.ai chat, and Claude Code differ and when to reach for each.
  • You've completed p10, p11, and p04.

See also