Your agent acts as your personal tutor, generating progressively harder problems. It tracks your skill level on each concept and picks the next problem
based on where you are.
Current courses: Algorithms & Data Structures and Concurrency.
Get started
Clone the project and cd into it:
git clone [email protected]:zuzuleinen/algotutor.git && cd algotutor
Run initial set-up to choose your agent and optionally enroll in a course:
make init
Train > Check > Review
Start a training session with make train.
Submit your solution by telling your tutor check. It's OK to say, "I don't know". They will pick an easier problem for you.
Start a daily review session with make review.
Read more here about the motivations behind the project.
Commands
Local commands — terminal
Command
What it does
make init
First-time setup — enroll in courses, pick a default agent
make enroll
Add another course to your enrollment
make train
Launch your agent in training mode for the active course
make train <slug>
Switch active course to <slug> and launch the agent there
make review
Open the review TUI across every enrolled course
make review <slug>
Open the review TUI scoped to one course
make list
List all available courses and your enrollment status
make run
Sanity-check your solution before check (active-course aware)
Agent commands — chat
Type these in your agent session:
Command
What it does
train
Get the next problem in the active course
train <course>
Switch to <course> and start training there
check
Submit your solution for evaluation (grading, mistake logging, level updates)
I don't know
Break the problem into simpler sub-problems
I want to solve [problem desc]
Request a specific problem with description or screenshot
reset
Wipe progress in the active course (with confirm reset gate)
reset all
Wipe progress in every enrolled course (with confirm reset all gate)
Requirements
An AI coding agent — see docs/agents.md for the supported list and setup
