ai-learning-lab

Universal sandbox for AI-assisted learning workflows and course experiments.

This repository is now learning-first. Previous Docker automation examples were intentionally retired so each new topic can be built in a dedicated course structure under learning/.

Why This Repo Exists

• Keep a lightweight place for practical AI-assisted learning.
• Compare workflows across tools with low operational overhead.
• Capture reusable patterns in one place without vendor lock-in.

Repository Structure

ai-learning-lab/
|- AGENTS.md
|- README.md
|- learning/
|  |- course-index.md
|  `- courses/
|- .github/
|  |- agents/
|  |- prompts/
|  |- instructions/
|  `- workflows/
|- scripts/
|  `- check_docs.py
|- tests/
|  |- unit/
|  `- integration/
|- requirements.txt
|- requirements-dev.txt
`- pyproject.toml

Quick Start

git clone https://github.com/tausagiz/ai-learning-lab.git
cd ai-learning-lab
python -m venv venv
source venv/bin/activate
pip install -e ".[dev]"

Working Model

• Use AGENTS.md as the central, vendor-neutral workflow policy.
• Keep adapter files in .github/agents/ and .github/prompts/ thin and focused.
• Add new course-specific code in isolated folders when needed, instead of growing one monolithic example.

Validation

Default low-cost checks:

python scripts/check_docs.py

Roadmap

Current Focus

• Establish baseline course skeleton templates in learning/courses/.

Next Up

• Add first universal course module independent of Docker/DevOps.
• Add a lightweight checklist for course structure quality.

Backlog

• Add optional examples for selected domains as separate course tracks.
• Add reusable course metadata conventions.

Done Recently

• Retired legacy Docker automation example to keep the repository learning-first.

License

MIT License - see LICENSE.

Copyright © 2026 Mateusz Rusnak.