Claude Code Day to Day: From First Prompt to Safe Habits

What's inside

1. Level 1: Agent, Not Chatbot: What Claude Code IsFree
   Learn what separates a coding agent that reads files and runs commands in your project from a chat assistant you copy and paste from, and what this independent course (not affiliated with or endorsed by Anthropic) will have you build on a small sample project.
2. Level 2: The Agentic Loop, Built-in Tools and ContextFree
   See how Claude Code works in a repeating cycle of gathering information, acting and checking results, using the built-in tools Read, Edit, Write, Bash, Grep and Glob, all inside a limited context window.
3. Level 3: Installing on macOS, Linux and Windows
   Install the Claude Code CLI with the method that fits your operating system, confirm it works and know how it stays up to date.
4. Level 4: Signing In: Subscription or Console Account
   Start your first session in the sample project and sign in with either a Claude subscription or an Anthropic Console account, understanding how each one is billed.
5. Level 5: One Engine, Many Surfaces
   Tour the terminal, VS Code, JetBrains, the desktop app, the web and the mobile app, and learn what stays the same across them and what differs.
6. Level 6: First Prompts on an Unfamiliar Codebase
   Use read-only questions to get Claude Code to map a project you have never seen before you ask it to change anything.
7. Level 7: Permission Prompts and Permission Modes
   Understand what Claude Code asks before doing, how to answer a permission prompt, and how the permission modes trade convenience against oversight.
8. Level 8: Plan Mode: Think Before You Touch Files
   Use plan mode to have Claude research the code and propose a written plan that you review and correct before any file is edited.
9. Level 9: Explore, Plan, Code, Commit
   Combine exploration, planning, implementation and verification into one repeatable workflow and take a small feature through it end to end.
10. Level 10: Git with Claude: Branches, Commits and PRs
    Let Claude Code handle branches, staging, commit messages and pull requests while you stay in control of what goes into history.
11. Level 11: Specific Prompts, File References and Images
    Get better results on the first attempt by naming files, constraints and success criteria, and by supplying files, pasted output and images as input.
12. Level 12: Managing Context: Compact, Clear, Rewind
    Inspect what is filling the context window, decide between compacting, clearing and starting a fresh session, and use checkpoints to rewind code or conversation when a turn goes wrong.
13. Level 13: CLAUDE.md, AGENTS.md and Auto Memory
    Give Claude lasting project instructions with CLAUDE.md, understand how user, project and nested files combine, and see how AGENTS.md and auto memory fit alongside it.
14. Level 14: Slash Commands and settings.json
    Drive a session with the built-in slash commands, and store permission rules and preferences in settings.json at the right scope.
15. Level 15: A First Skill and a Subagent Code Review
    Package a repeatable instruction as a small skill, then create a subagent with its own context window and use it to review your changes.
16. Level 16: Connecting an MCP Server and a Format Hook
    Connect one external tool through the Model Context Protocol, and add one hook that runs a formatter automatically after Claude edits a file.
17. Level 17: Cost, Usage and Safe Review Habits
    Track what a session costs or how much of your plan it uses, and build the habits that stop leaked secrets, unreviewed diffs and destructive commands.
18. Level 18: Final Challenge: Ship a Feature End to End Exam
    Review the whole course by taking one feature on the sample project from first exploration to a reviewed pull request, using every habit and tool covered.

Access

The first 2 levels are free with a free account. Every level, the podcast edition and the AI tutor come with All Access at £4.99/month or any Creator plan — see pricing.

Start free →