Stanford course teaches developers to build agent-ready codebases
Stanford adjunct professor Mihail Eric is teaching a 10-week course that prioritizes context engineering and security over manual coding. The curriculum reflects a shift where developers use coding agents exclusively and must design systems for autonomous verification.
Mihail Eric, adjunct professor at Stanford and head of AI at Monaco, is teaching CS146S, 'The Modern Software Developer,' for the second time, having first taught it in fall 2025. The 10-week curriculum targets experienced developers with prerequisites in algorithms, focusing on agent-ready codebases, context engineering, and security. In the first week, students build a coding agent harness in 200 lines of code and use an MITM proxy to inspect Claude Code's system prompt, which Eric noted was reduced by 80% at one point but has since grown again.
The course emphasizes verifiability and service-level boundaries, with week seven dedicated to security and new attack vectors from autonomous systems. Eric argues that agent-ready codebases are a prerequisite for multi-agent systems and software factories. Students are no longer typing by hand and are using coding agents exclusively, shifting their focus to managing agent workflows. The curriculum includes open-source contributions with partners like Factory, Cognition, and OpenAI, where students work on complex codebases and interact with maintainers. Eric maintains that foundational software engineering knowledge remains essential, but the focus must shift to context engineering and architecture to be productive in the modern workforce.