Course Summary: What You Now Know

You have finished the Memory System course. Here is everything you now know and why it changes everything that comes after this.

What you learned

Why Claude forgets you: every new Claude Code session starts completely blank. Without a memory file, Claude does not know your project, your preferences, or anything you have told it before. This costs time every session, forever. CLAUDE.md solves this permanently.

What CLAUDE.md is: a plain text file named exactly CLAUDE.md where the uppercase letters matter. It lives in your project folder and Claude reads it automatically at the start of every session before you type a single word. It gets loaded into Claude's context as if you had typed all of it yourself.

The two levels: your global CLAUDE.md stored at ~/.claude/CLAUDE.md holds your personal preferences and applies to every project. Your project CLAUDE.md stored in the project root holds project-specific context. Both are read every session and they combine rather than cancel each other out.

How to create one: two methods. Run /init in an existing project and let Claude generate a starter, or create the file manually. After creation, run /memory in any session to confirm Claude is reading it correctly.

What to put in it, the six sections:

• Section 1 is about the project: what it is, who it is for, and what stage it is at.
• Section 2 is what has already been built: so Claude does not rebuild or accidentally break existing work.
• Section 3 is tech and tools: exactly what you are using and just as importantly, what you are not.
• Section 4 is preferences and style: how you like things to look and how Claude should communicate with you.
• Section 5 is how to work with you: when to ask, when to act, and how much autonomy Claude should take.
• Section 6 is boundaries: what Claude must never do without your explicit permission.

How to build it over time: start with the six sections and grow through friction. Every time you have to repeat yourself, every time Claude makes the same mistake twice, every time you fix a painful problem, add a line. At the end of any difficult session, ask Claude what it would add to your CLAUDE.md based on what happened today.

Sub-agents: specialist assistants defined in a folder called .claude/agents/ inside your project or home directory. Each has a specific role and hard limits on what it can touch. Use /agents to create them. Most useful for larger, multi-area projects. Start with CLAUDE.md alone and move to sub-agents when your projects grow to need them.

What you built

A CLAUDE.md file with all six sections filled in. A global preferences file for your personal settings. And the habit of growing both through real use.

What changes now

Every future Claude Code session in a project with CLAUDE.md will start with Claude already knowing what you are building and who it is for, what is already done, what tools you are using, how you like things, and what Claude should never do.

The setup you did in this course will compound. The more sessions you run, the more refined your CLAUDE.md gets and the better Claude performs, without you having to explain anything from scratch.

What is coming next

Track 2: Websites

Now that Claude knows who you are and how you work, you are ready to build your first real website. One you could actually show to a client or a potential employer.

Track 2 covers the full website build: structure, design, multiple pages, mobile responsiveness, and deployment. By the end you will have a live website built entirely through conversation.

Before you move on

The homework for this course is in the next lesson. It takes 20 minutes and cements everything you just learned. Do not skip it.