Level 0 · 8 lessons · Start here
The Claude Interface
The conversation window has depth most users never find.
Before any Projects, Connectors, or automation, the conversation interface itself has depth most users never find. This module covers what Claude actually is, how to read its confidence, when to switch models, and how to format output for what you actually need. By Lesson 8 you understand the free versus Pro tier and what upgrading unlocks.
Worth knowing:Claude does not know when it is wrong. Recognizing the signals that warrant verification (dates, statistics, quotes, technical claims) is the most important critical thinking skill for any Claude user.
Curious non-developersFirst-time Claude usersAnyone treating Claude like Google
Start Module 1 →Levels 1–2 · 8 lessons
Projects + Memory
The two systems that fix Claude's default amnesia.
Claude's default state is amnesia, and Projects and Memory are the two systems that fix it. Projects give Claude your standing context through Instructions and Knowledge Files. Memory stores specific facts Claude learns across all conversations. Used together, they eliminate 90% of the repetition that makes Claude feel like starting over every time.
Worth knowing:Projects and Memory are commonly conflated. Projects persist context per workspace; Memory stores facts across all conversations. Confusing the two causes most of the 'why does Claude keep forgetting me' frustration.
Frequent Claude usersPeople rebuilding context every chatPrivacy-conscious users
Start Module 2 →Levels 1–3 · 8 lessons
Artifacts
Claude produces things you use, share, and build on.
Claude does not just produce text. It produces things: documents, interactive web apps, diagrams, and tools you can use directly. This module covers all 7 Artifact types (Text, Code, HTML, SVG, Mermaid, React) plus how to iterate, refine, and extend each one. By the end you can ship a mortgage calculator, a flowchart, or a habit tracker from a single conversation.
Worth knowing:An HTML Artifact for a mortgage calculator is not a document you save. It is a running web app you use. Most users miss this distinction and treat Artifacts as fancy file downloads.
Creators who want working toolsDesigners and PMsAnyone who saw ChatGPT canvas and wanted more
Start Module 3 →Levels 2–4 · 8 lessons
Connectors + Routines
Claude works with your real data, on a schedule.
Connectors give Claude read (and sometimes write) access to your actual data: Google Drive files, calendar, email, project tools. Routines run tasks on a schedule you define without manual triggers. This module covers the 375+ Connector library, the read versus write distinction, and how to build your first automated weekly briefing in 15 minutes.
Worth knowing:Not all Connectors are read-only. Some can write back, posting to a tool, updating records, or sending messages. Knowing which is which determines what Claude can do autonomously.
Pro users on Google WorkspaceOperations and project managersAnyone with weekly recurring work
Start Module 4 →Levels 3–5 · 10 lessons
Desktop Deep-Dive
Cowork, Computer Use, and Chrome.
The Desktop app is not just Claude installed on your computer. It has a mode (Cowork) that runs scheduled work, controls software via Computer Use, and persists session state. This module covers Chat versus Cowork contexts, Computer Use sessions, the Chrome extension, and combining scheduled tasks with Computer Use for autonomous workflows.
Worth knowing:Most users never discover Cowork. The default Chat experience in the Desktop app reveals nothing about scheduled tasks or Computer Use. That is exactly why this track exists.
Mac and Windows Desktop usersPeople with repetitive computer tasksCowork track prerequisites
Start Module 5 →Levels 4–5 · 8 lessons
Claude Code Surfaces
What "Claude Code" actually means (three products).
Claude Code refers to three different products: the Code tab in claude.ai, the Claude Code desktop GUI, and the Claude Code CLI for developers. Most non-developer confusion about the platform starts here. This module disambiguates all three, covers GitHub for non-developers, and identifies the signals that mean you are ready for the Code track.
Worth knowing:Non-developers do not need the Claude Code CLI. The Code tab in claude.ai is plenty for editing documents, working with structured data, and reviewing outputs without writing a line of code.
Non-devs on engineering teamsPMs reading pull requestsAnyone confused by 'Claude Code'
Start Module 6 →Levels 3–5 · 7 lessons
Platform Mastery
Skills, billing, a personal Claude OS, the track map.
The last gaps in the mental model: Claude Skills (pre-built workflows), model selection as a judgment call, what your plan tier actually includes, and how to build a Personal Claude OS. The module ends with the canonical track map: which vertical track to take next based on the work you actually do.
Worth knowing:Pro includes Projects, Connectors, Routines, Computer Use, and longer context. The upgrade decision is simple: if you want Claude to remember context or connect to your tools, you need Pro.
Pro subscribersTeam leadsAnyone planning their next track
Start Module 7 →