Understand what you're exporting
Interactive lessons on how AI tools actually work, context, instructions, projects, and agents. No jargon, no fluff. Each lesson ends with a quiz.
Foundations
Understand how AI chat works — context, instructions, and why setup matters.
How AI chat works
Each reply is predicted from the full conversation so far, the model does not reason, and without memory features it starts completely fresh in every new chat.
The context window
The context window is how much text the model can see at once, fill it and quality degrades, so start a fresh chat when it runs low.
Instructions vs messages
Instructions are persistent, they set the role, tone, and rules for every reply, so you never need to repeat yourself.
ChatGPT
Set up custom instructions, projects, and knowledge files to get consistent results.
Custom instructions
Tell ChatGPT who you are and how you want it to respond once, every new chat picks it up automatically, no re-typing needed.
ChatGPT Projects
A ChatGPT Project groups related chats under shared instructions and files, so every conversation starts with consistent context, no re-pasting.
How file uploads work
Larger files are indexed, not read in full, the model sees only the most relevant passages, so specific questions get far better results than vague ones.
GPTs vs Projects
A custom GPT is a shareable assistant you publish for others; a Project is a personal workspace for your own ongoing work.
Claude App
Use Claude Projects and styles to turn the app into a tuned assistant for your work.
Claude Projects
A Claude Project pairs persistent instructions with knowledge files, every chat inside the Project inherits both automatically.
Claude Styles
Claude Styles set tone, length, and formality across all your chats at once, use a preset or describe your own in plain English.
Claude Artifacts
Ask Claude for a standalone deliverable and it opens in a side panel, ask for changes and Claude updates it in place.
Claude vs ChatGPT
Claude and ChatGPT share the same core concepts but use different names, this lesson maps them side by side, including gaps where one has no equivalent.
Claude Code
Understand agents, skills, harnesses, and the loop that powers agentic coding.
From chat to agent
Claude Code is an agent: it reads files, runs commands, and checks its own work, then keeps going until the task is done.
Reusable skills
Write a procedure once as a skill and Claude Code loads it automatically whenever your request matches, no slash command, same steps every time.
Agents & subagents
Claude can delegate large or independent tasks to a subagent in its own context window, getting back only a summary, the main conversation stays compact.
The harness
The model is one part of Claude Code; around it is a harness of tools, permission rules, CLAUDE.md, and event hooks. You configure the harness, not the model.
AI Engineering
The three levers behind good AI output, prompt, context, and loop engineering, and how to pull each one.
The three levers
Better prompts are just one of three levers, what the model can see and what happens after its reply often matter more.
Prompt engineering
The words you choose are the most visible lever on AI output, and often the least powerful one, so use them well.
Context engineering
Controlling what the model can see when it answers often matters more than how you phrase the question.
Loop engineering
The third lever is not what you say or what the model sees, it is what happens around each response, and it is where most silent failures get caught or missed.