A tiny reference repo for confirming a developer's environment is ready before day one. It exercises the whole Docker + uv + Git/GitHub toolchain end to end, so a single git clone and four commands tell you whether someone is good to go.

There are two layers to validating a setup:

1. Quick per-tool checks — fast commands anyone can paste into a terminal.
2. End-to-end smoke test — clone this repo and run four commands that confirm the whole chain works together.

📋 This repo covers the tooling prerequisites. For the full picture — required setup, prerequisite knowledge, and refresher tracks — see the Agentic AI Academy prep package.

Run each block in a terminal. Every command should succeed without an error.

# Docker — installed AND the daemon is actually running
docker --version
docker run --rm hello-world

# uv — installed and able to manage Python
uv --version
uv python install 3.11
uv python list                 # should show 3.11+

# Git + GitHub access (auth is the usual failure point, not git itself)
git --version
ssh -T git@github.com          # SSH: expect a "successfully authenticated" message

# Azure CLI — just needs to be *installed*; account access comes on day 1
az --version

No need for az sign-in check Azure account access is handed out on day 1, so don't expect to be signed in yet — having the az CLI installed is all you need beforehand.

Clone this repo and run the four commands below. Green on all four = environment confirmed.

git clone git@github.com:datarootsio/a3-setup-check.git && cd a3-setup-check
uv sync                          # uv resolves + installs into the right Python
uv run pytest -q                 # runs the suite locally
docker build -t setup-check .    # confirms you can *build* images
docker run --rm setup-check      # runs the same tests inside the container

Why four separate commands? Each catches a distinct failure mode:

It's intentionally tiny — three files do all the work:

setup-check/
├── pyproject.toml      # project + a single dev dependency (pytest)
├── Dockerfile          # python:3.11-slim + uv, runs the tests as its CMD
└── tests/
    └── test_smoke.py   # asserts Python >= 3.11

The Dockerfile pulls the uv/uvx binaries straight from the official ghcr.io/astral-sh/uv image, installs dependencies in a cached layer, and runs uv run pytest -q as its default command — so building and running the image re-validates the exact same flow you ran locally.

A couple of choices are baked into the checks above. Swap them if your environment differs:

• uv-managed vs. system Python. By default we let uv own the 3.11 runtime (uv python install 3.11) — the cleaner, most reproducible path. If you must use a system-installed Python instead, replace that line with python3.11 --version and add uv venv --python 3.11.
• SSH vs. HTTPS for GitHub. The quick check uses SSH (ssh -T git@github.com). If you authenticate over HTTPS + a personal access token (PAT) instead, drop the ssh -T line and do a test clone of a target repo — the auth mechanism is different.