Session 00 — Get ready to learn

Time: 120 minutes with your instructor. Keep this checklist open. Use your own accounts and a local course copy.

Quick start

First action: Verify Node, open the course folder and start the local practice app.

Finish with: A private readiness card and one checked AI explanation.

Step-by-step help for Session 00 — smaller steps, checkpoints and troubleshooting. Saving and asking for help.

Mac / Linux / Windows

Follow the platform guide for your OS and shell. The core course does not require WSL, Homebrew, Docker or desktop Office. Record your OS, shell and browser on your readiness card; ask a coach before changing managed-device settings.

1. Create your GitHub account

Open https://github.com/signup. Choose a username and complete email verification. Existing account? Sign in at https://github.com/login instead of making a duplicate.

Open https://github.com/settings/security to configure two-factor authentication. Store recovery codes privately. Review https://github.com/settings/emails before publishing commits so you understand email visibility.

GitHub's account guide: https://docs.github.com/en/account-and-profile/how-tos/account-management/creating-an-account-on-github

2. Find the course repository

Optional: open My learning path in the course navigation. The planner guide explains private browser-local progress, evidence references, hints and backup/restore. Do not enter secrets or personal records; your checkmarks do not award a certificate. Keep an exported backup if you use it.

Use Course repository in the site navigation, or the exact URL your instructor supplies. Confirm the name is ai-quality-academy. Read README, open labs/, and find 01-ai-claim-detective.md.

Public reading does not require membership in the owner's organization. A star bookmarks the project; it does not give write access. A fork is your copy on GitHub; a clone is a local Git copy; Download ZIP is a local file copy without Git history.

3. Install the local tools

Node.js: https://nodejs.org/en/download — install a supported LTS version, version 22 or newer. On Windows, use the prebuilt Windows Installer (.msi) download. Chocolatey and optional native-build tools are not required for this course. If a package-manager installation gets stuck, use the setup walkthrough to verify Node and choose the direct installer route with a coach. Optional GitHub Desktop (macOS/Windows only): https://desktop.github.com/download/ Linux: use ZIP or optional Git CLI from https://git-scm.com/downloads. Optional editor: https://code.visualstudio.com/download

Reopen Terminal (macOS/Linux) or Command Prompt/PowerShell (Windows). Check:

node --version
npm --version

If an organization-managed device prevents installation, ask your instructor. You can read the site and pair on an approved machine while access is arranged.

4. Get a local copy

Simple path: repository → Code → Download ZIP → extract → open the extracted folder. Git path: GitHub Desktop → File → Clone repository → URL → paste the exact course URL → choose a local folder.

GitHub Desktop is for Mac/Windows. On any OS with Git installed, run git clone https://github.com/labalicious-learning/ai-quality-academy.git, then open the cloned folder. ZIP users on Windows must choose Extract All before opening it. An editor's File → Open Folder followed by Terminal → New Terminal avoids platform-specific path commands.

Open a terminal inside that folder. Confirm package.json is present. You do not need admin rights to the course repository.

5. Run your first lab

npm start

Open http://127.0.0.1:4178. Leave the terminal running. Select Candidate B, click Load allocation and confirm a reserved record appears. This is a startup check, not an investigation of the teaching defects. Nothing is sent to a real business.

Press Ctrl+C in the terminal to stop. Run npm start again to restart. On Mac this is Control+C, not Command+C. In Windows PowerShell, use npm.cmd if script policy blocks npm; do not weaken execution policy. Command Prompt is also supported. For built-in checks, stop the server if desired and run:

npm test
npm run verify

The basic lab uses Node's built-in libraries. npm ci is only needed for rebuilding the website/slides.

6. Set up Codex or an approved compatible client

Follow the current official quickstart: https://developers.openai.com/codex/quickstart

Use the client/host route approved for your OS and budget. Codex remains the main teaching client; the small-model guide also covers Qwen3.8-27B through OpenCode, local inference and verified custom-provider routes. No premium subscription is mandatory. A hosted API route may need a dedicated capped key; local inference need not. Sign in or connect through the client, never through a course file. Your instructor confirms actual model/tool compatibility before Session 02.

Open your local course folder. Ask:

Read only fixtures/setup-example.mjs. Explain describeCount in plain language without changing files. Predict its output for 1 and 3, cite the relevant lines, and tell me how to check your claim.

Open the cited function yourself, then run node fixtures/setup-example.mjs and verify the prediction. This unrelated practice file does not reveal later lab solutions. Record the model/client shown in your account; do not guess based on the course date.

7. Save your readiness card

This card stays private. Show readiness to your instructor without exposing account or security details. Later labs use the separate submission workflow; do not upload this setup card there.

Create a personal note outside the public repo, or in the ignored submissions/ folder:

• GitHub account verified: yes / blocked
• OS/version, shell, browser/version:
• Course repo located: yes / blocked
• Local lab starts and stops: yes / blocked
• Approved AI client/model can explain a file: yes / blocked
• One claim I checked:
• One question for my instructor:

Do not record passwords, email addresses, tokens or recovery codes.

If something fails

• “npm not found”: reopen terminal; check Node installation.
• “package.json not found”: open the extracted course folder, not its parent.
• “Address already in use”: stop your earlier course server or use the platform guide's per-shell 4179 instructions with a coach. Do not kill unrelated programs.
• Browser cannot reach localhost: keep npm start running.
• GitHub works but Codex does not: they have separate logins and entitlements.
• No AI access yet: finish setup and evidence reading with a partner. Do not buy an unapproved subscription to finish class.

Ready for Session 01

You can start the lab, find the next handout, and explain where the files live.

Before Session 01: a product you might enjoy building

Browse Product Studio, the idea bank and small-model setup options. You do not need to buy a subscription or create the product repo today. Bring an idea to Session 01; instructor approval and kickoff happen by Session 02. Arrange help privately if accounts, hardware or cost block your route.