Course Description
Large language models (LLMs) have transformed software development from a purely hands-on coding process into one carried out alongside increasingly capable coding agents. This shift requires rethinking how intent is defined, how work is structured, and how tools are coordinated so that agents can contribute effectively to complex software projects.
This course covers the current practices and technologies behind AI-native software development, including MCP (Model Context Protocol), agent skills, spec-driven development, context engineering, and the software factory. Students will learn how to provide agents with the right context and capabilities, how to translate product requirements into executable specifications, and how to design iterative workflows in which humans and agents jointly plan, build, evaluate, and improve software.
Through hands-on assignments, a group project, and weekly sessions, students will explore both the capabilities and the limitations of today's coding agents. By the end of the course, students will be able to design effective agent-driven workflows, turn tools and skills into reliable development systems, and apply software-factory principles to build software faster and at greater scale.
Course Details
Tools and Platforms Used
The agentic development ecosystem used throughout the course:
- Antigravity — main development/experimentation environment
- Google ADK (Agent Development Kit) — for agent integrations
- 5 students per project group, working from a shared Google Cloud workspace
Classroom Management
Each week opens with a recap of that week's agenda, deadlines, and group progress. Attendance and participation are taken at the start of class; the last 10–15 minutes of each session are reserved for questions and discussion.