
Harness engineering: leveraging Codex in an agent-first world
By Ryan Lopopolo, Member of the Technical Staff
Project-based course on environment design, state management, verification, and control systems for AI coding agents.
Built on OpenAI and Anthropic documentation, it teaches harness design to constrain and verify long-running agentic workflows. Includes lectures, hands-on projects, and copy-ready templates like AGENTS.md and feature_list.json.