Agentic AI Coding with Claude Code¶
Welcome to Claude Code LabCamp. In this hands-on session, you will learn Claude Code by building a live trivia quiz app together, then testing it with the whole room. Plan for about two hours for the core labs and final project (including setup and breaks).
Documentation: storm-ai-reply.github.io/ClaudeCode-Labcamp
Repository: github.com/Storm-AI-Reply/ClaudeCode-Labcamp
How it works¶
Each participant works on their own laptop with personal, temporary AWS Bedrock credentials (provided by the organizers). You will go through four short labs individually, each teaching a Claude Code feature. Then you form teams of 5 for the final project: collaborate, split the work, and ship the most beautiful quiz in the room.
Getting started¶
-
Step 1: Set up your environment
Follow the setup instructions for Python, Git, Claude Code, and the project dependencies.
-
Step 2: Start Lab 01
Open Lab 01: The Execution Model and follow the exercises in order, or open the full lab map first.
Labs at a glance¶
Each lab is split into separate pages (overview → exercises → wrap-up) so you can focus on one step at a time. Use the Labs tab and sidebar for navigation; All labs (map) lists everything on one screen.
| Lab | ~Time | |
|---|---|---|
| 1 | The Execution Model | 30 min |
| 2 | Project Configuration | 25 min |
| 3 | Control & Connect | 30 min |
| 4 | Scale & Reuse | 25 min |
| Final Project | 30 min |
Work through them in order. Each lab has its own starter code. If you break something, run git checkout -- . to reset.
The competition¶
At the end, teams of 5 host their quiz. The room joins via QR code and plays a round. Best quiz wins.
| Criterion | Points |
|---|---|
| Works end-to-end (join, answer, leaderboard) | 25 |
| Claude Code setup (CLAUDE.md, commands, hooks, skills, MCP, subagents) | 25 |
| Quiz content (creative, entertaining, on-theme) | 25 |
| Beautiful UI (visual craft, polish, identity) | 25 |