AI Agent Guidelines for CS336 at Stanford
Stanford CS336 defines AI assistants as teaching aids, not assignment solution generators.
Stanford CS336’s CLAUDE.md sets boundaries for AI coding assistants such as ChatGPT, Claude Code, GitHub Copilot, and Cursor. Agents may explain concepts, review student-written code, suggest debugging checks, and point to course materials. They should not write code, complete TODOs, edit repositories, run shell commands, or implement core assignment components for students.
Stanford CS336 added a document called CLAUDE.md — an "AI Agent Guidelines" — to its assignment1-basics repo, with the goal of governing how students interact with AI coding assistants when using them. The document explicitly states that these AI tools include ChatGPT, Claude Code, GitHub Copilot, Cursor, and others, but that their role in the course should be "Teaching Assistant," not "Solution Generator." In other words, AI may help students understand concepts, clarify error messages, check completed code, and suggest edge cases, invariants, sanity checks, toy examples, assertions, profiling, and debugging methods; but it should not directly complete the implementation for students.
Free shows the 3-line summary; Pro unlocks the full deep summary (~300 words) so you never have to click through.
See Pro plans →Want the original English / full article?
Read on Hacker News (AI keywords) →Related
Summaries are AI-generated; the original article is authoritative.