Hands-on exercises1 scaffolded lab
Production-Ready Systems with LLMs and Agents: exercises
The labs are TypeScript, and every model call goes through one module so the provider stays swappable. The workflow router lab ships a starter, a worked solution, and a test suite that fails on a fresh clone: the tests are the brief. The remaining labs are specified end to end inside their lessons, with the steps, the acceptance criteria and the code shape, and you build them in your own codebase.
How the exercises work
A lab ships a starter you fill in, a worked solution published alongside it, and a test file that fails on a fresh clone. The test file is the brief: read it first, then make it pass. Every model call goes through common/llm.ts, so swapping provider or model is a one-line change and the lab still teaches the same thing.
git clone https://github.com/ehsangazar/llms-and-agents-labs.git
npm install
cp .env.example .env # add your OPENROUTER_API_KEY
npm test # the tests are the brief; they fail on a fresh clone
npm run lab modules/01-workflows-and-agents/lab-workflow-router/starter/index.ts