Week 1

📄 Optional Homework

Resources

Notes

Channels

Messages

Channels

Messages

Back

Lesson 1: Fundamentals & Lifecycle of Application-Centric Evals

📄

Optional Homework

Lesson 8

Lesson 1’s homework is about building and prompting a Recipe Suggestion Bot. The bot’s job is simple: given some ingredients or constraints from a user, it should suggest recipes that are safe, practical, and easy to follow. Think of it like a cooking assistant powered by an LLM—it takes user input (e.g., “I have chicken, rice, and broccoli”) and returns a recipe idea with clear steps. Your task is to design the system prompt that makes the LLM behave like this assistant: define its role, decide what it must always do (give valid recipes, include ingredients, stay safe) and what it must never do (unsafe food advice, irrelevant chatter), how much creative freedom it has (stick to exact inputs or allow substitutions), and how the output should be formatted.

To get started, you’ll need:

If you’re new to programming, check out the #beginner-programming channel in Discord for help, including AI-assisted programming tips. Additionally, for those new to coding, Isaac Flath will be doing a three part mini-course during weeks 3-4 titled "Live coding workshop ... w/ Isaac Flath" for each of the homework assignments, where he will step through how to complete the homework assignments with AI assistance.

Repo link for instructions and scaffold code: hw1.

Remember, the homework is totally optional! It is meant for folks who want to get lots of hands-on practice. We estimate that this homework can take between 1 and 3 hours for folks who already know programming. You do not have to do the homework to get a lot of value out of the course.

[

Previous

](/parlance-labs/evals/2025-3/syllabus/modules/98b8aa?item=7aduv0awiub)

CompleteComplete this lesson

75%

[

Home

](/parlance-labs/evals/2025-3/home)[

Community

](/parlance-labs/evals/2025-3)