Assignment 3: Build a Neat Demo
Overview
Use vibe coding to build something cool! This assignment is your
chance to explore the AI-assisted development workflow and create a
working demo that showcases your creativity and technical curiosity.
Your demo can be anything — an interactive tool, a data visualization,
a game, a useful utility, a web app, or something entirely different.
The only requirement is that you build it using vibe coding techniques
(AI coding agents like Claude Code, chat.dartmouth.edu, or similar tools).
Deliverables
1. Working Demo
A functional project that does something interesting. It can be:
- A single file (HTML, Python script, Jupyter notebook)
- A multi-file project (web app, CLI tool, etc.)
- A Colab notebook with interactive elements
Push your code to your GitHub repository for this assignment.
2. YouTube Video (~5 minutes)
Record a video that tells the story of your workflow — not just a
demo of the final product. Your video should include:
- What you built and why (brief — ~1 minute)
- Your workflow: Show screen recordings of how you used AI tools.
- Tips and tricks: Share specific, evidence-based examples of what
What prompts did you use? How did you iterate? (main focus — ~3 minutes)
worked well and what didn't. Show actual prompts, AI responses, and
your iteration process. (~1 minute)
The goal is to help your classmates learn from your experience. Think
of it as a tutorial on your vibe coding workflow.
Example Project Ideas
Looking for inspiration? Here are some ideas spanning different domains:
- Interactive data dashboard: Load a public dataset and create an
- A simple game: Build a browser-based game (puzzle, quiz,
- Useful CLI tool: A command-line utility that solves a real problem
- Data story visualization: Find an interesting dataset and create
- Web app: A single-page app that does something fun or useful
interactive visualization with filters and charts
platformer) using HTML/CSS/JavaScript
(file organizer, text summarizer, study timer)
a compelling visual narrative
(weather dashboard, recipe finder, habit tracker)
These are just starting points — feel free to go in a completely
different direction!
Evaluation Criteria
Your work will be evaluated on:
| Criterion | What we're looking for |
|---|---|
| Creativity | Is the demo interesting, surprising, or useful? |
| Workflow clarity | Can we follow your development process from the video? |
| Evidence quality | Do you show specific prompts, iterations, and debugging? |
| Understanding | Can you explain what the code does and why it works? |
| Storytelling | Is the video engaging and well-organized? |
Note: We are not evaluating code complexity or polish. A simple but
well-understood demo with a great workflow video scores higher than a
complex project you can't explain.
Verify and Explain
Before submitting, make sure you can answer these questions:
- What does each major section of your code do?
- Why did you make the design choices you made?
- What would you do differently next time?
- What was the most effective prompting strategy you discovered?
Resources
- chat.dartmouth.edu — Free access to Claude, ChatGPT, Mistral, and more
- claude.dartmouth.edu — Dedicated Claude access
- Claude Code documentation — CLI coding agent
- GitHub Student Developer Pack — Free GitHub Copilot and more
- Google Colab — Browser-based Python notebooks with AI features
Submission
- Push your demo code to your GitHub assignment repository
- Upload your YouTube video (unlisted is fine)
- Submit the YouTube link via the course submission form