TIP
I’m developing a new course for S2 2026: COMP4020/8020 Rapid Prototyping for the Web. This post is me thinking aloud as I work through the course design—nothing here is final, and feedback is welcome.
Course Description (aka “the pitch”)
LLM agents, which run tools (e.g. search the web, edit files, run code) in a loop to achieve a goal, offer a new workflow for developing software. This is especially true for application domains which are well-represented in LLM training sets—like the web.
Moving beyond naive “vibecoding”, this course provides a studio-based “iterate and test with working prototypes” approach to rapidly prototyping web apps. In the weekly lectures students will learn fundamental concepts and practical skills for harnessing an agentic AI/LLM development workflow. In the weekly “studio session” tutorials, students will demo their work-in-progress prototypes and receive feedback from peers and instructors. By the end of the course each student will have designed, developed and deployed multiple web app prototypes using a rapid, feedback-driven process.
What’s the background to this course?
Most people point to the ChatGPT release in November 2022 as an inflection point in the use and influence of LLMs in modern life, but for software developers it’s Claude Code’s release in May 2025 which has made the biggest change to the way we do (some of) our job. I don’t subscribe to either the “booster” or “doomer” end of the spectrum with AI/LLM tools; but I am curious about the way they’ll change the way that people make cool stuff.
Anyway, the ANU School of Computing (where I was formerly a faculty member, before moving to the School of Cybernetics in 2021) has asked me to create a new course for 2026 and this is the course I think we need right now. It’s the course I want to create, anyway.
Draft learning outcomes
INFO
These are draft learning outcomes—subject to change as the course develops.
At the completion of this course, students will be able to:
- design, build and test full-stack web applications using a rapid-prototyping process
- describe the components of a Large Language Model interface for code generation
- design and evaluate different LLM agent workflows for software development
- apply principles from the scholarly literature to work-in-progress and finalised software prototypes
What’s next?
I’m still working through the assessment structure, the weekly schedule, and the technical infrastructure. If you’ve got thoughts on teaching agentic software development—or you’re doing something similar at your institution—I’d love to hear from you. If you’re a student at the ANU School of Computing, and have a slot for a 4XXX/8XXX elective in S2 2026, then keep an eye out for more information as the time approaches.