Course 1 · Understand

How AI agents actually work.

Thirteen short lessons on what is really going on inside an AI agent: the model, the prompt, tools, context, memory, the loop, evals and permissions. No jargon left unexplained, and it works whichever tool you use.

FreeBeginner13 lessons · 98 minWorks in any AI tool
Start lesson 1No signup to read.
Joe Leung

Written by Joe LeungCTO and co-founder, LaunchLemonade

13 lessons · about 98 minutes

Course outline

Best in order, but every lesson stands on its own. Jump to the one you need.

  1. 1What is an LLM?What a large language model really does, why it can seem so clever, and what it can't do on its own.7 min
  2. 2Prompts: brief it like a contractorWhat a prompt really is, why it isn't dead, and how to write a clear brief that is neither too vague nor too specific.8 min
  3. 3Tools: how an AI does thingsHow a model that only writes text can check the weather, read your files or send an email, and why some tools need a human to say yes.8 min
  4. 4Context: what's on the deskWhy more information isn't always better, and how to decide what an AI should see right now and what it can safely ignore.8 min
  5. 5Memory: the filing cabinetHow an agent carries useful facts from one conversation to the next, and why remembering everything is as bad as remembering nothing.7 min
  6. 6What is an AI agent?An agent isn't a smarter model. It's a model plus a harness, going around a loop until the job is done.8 min
  7. 7Evals: how do you know it works?Why "I tried it a few times and it seemed fine" isn't enough, and the four layers of checks that tell you an agent actually works.8 min
  8. 8Who approves what an agent does?How to decide which actions an agent can take on its own, which need a human, and why some risks can't be fixed with better wording.8 min
  9. 9RAG: finding the right pageHow an agent answers from your documents without reading all of them, and the three ways it goes wrong.7 min
  10. 10Subagents: when an agent delegatesWhy serious agent systems hand jobs to other agents, and how to write the brief, shape what comes back and limit what each one can do.7 min
  11. 11MCP: one plug for every toolWhat the Model Context Protocol is, why it changed how agents reach your apps, and the questions to ask before you connect one.7 min
  12. 12Not everything needs to be an agentThe difference between a workflow and an agent, why autonomy is a dial rather than a switch, and where to start.7 min
  13. 13The whole pictureEvery part of an agent on one page, with one job running through it, and the questions to ask before you build or buy one.8 min

After this course

Get better answers from your AI assistant

Five practical lessons for when your assistant is nearly right: what to check when it goes wrong, how to write its instructions, what to give it to read, and how to tell whether it's actually working.

See the course