Subjects › AGI Agents (6-week course)
W0 · Setup, tests & model backends
How the course repo runs: baseline vs exercise tests, CI, the offline mock model, .env settings and the Python version pin.
The idea
Two test suites with opposite jobs: `pytest -m "not exercise"` is the baseline and must always be green (CI enforces it); `pytest -m exercise` starts red and is your to-do list (CI reports it but never fails on it). Everything runs offline on a deterministic MockLLM unless AGI_LLM_PROVIDER (in your shell or .env) says otherwise — the test suite always forces mock via conftest.py; real environment variables beat .env, and within .env the last occurrence of a key wins. Use Python 3.12 or 3.13 — CrewAI 1.x does not support 3.14. In your repo (tag v1.0.1): Week00_Setup/README.md, check_env.py, agi_core/config.py, pyproject.toml, the Makefile and .github/workflows/ci.yml.
Lesson
No lesson built yet. The worked solutions below are the material it would be assembled from.
Questions (5)
| Question | Type | Difficulty | Drift | Accuracy |
|---|---|---|---|---|
| You've just cloned the course repo and installed the requirements. Which command should come back all green be… | mcq | Easy -1.40 | — | — (0) |
| You push on Wednesday of Week 3 with exercises 3.4 and 3.5 still unfinished. The CI workflow has a `baseline` … | mcq | Easy -0.60 | — | — (0) |
| Your laptop's default Python is 3.14. You create a venv with it and, before cloning the course, run a bare `pi… | mcq | Moderate 0.20 | — | — (0) |
| Your `.env` contains, in this order: AGI_MODEL=qwen2.5-coder:14b AGI_MODEL=llama3.1:8b and you have … | mcq | Hard 1.00 | — | — (0) |
| The v1.0 loader stored everything after the first `=` as the value, trimming only surrounding spaces and quote… | short | Very hard 1.50 | — | — (0) |