It is clear that LLMs have made learning much easier, but they have the potential to be great teachers if it has the correct context on the user's understanding of the topic. This repository aims to give your harness that exact context, inspired by Eero Alvar's video How I Use AI to Learn Things (https://www.youtube.com/watch?v=kzcI5F4tGiU). This repo has just adapted his setup a bit but serves the same function of:

1. Testing you: to see how much you know, so it can adapt the content of the curriculum.
2. Outlining an exact plan for you to know how the LLM is going to be teaching you.
3. Teaching you the curriculum with short quizzes to see how much you have truly understood.

setup

I am using my learning Obsidian vault as the home of the three folders because:

• it gives the harness more context about how much I already know in regards to the requested topic
• it will also create a new markdown file to display the questions and teaching material, so you can easily review it again
• Obsidian will also render the equations, so it is easier to understand if you are using it to learn STEM subjects
• it also saves all the progress you have made in each context window so the agent would adapt to your learning style in the next session.

To run it, there is nothing special but to just copy these three folders into your main folder, and run the harness on the terminal of the overarching folder.