Agentic Engine Optimization (AEO) (addyosmani.com)

A new concept called Agentic Engine Optimization (AEO) has emerged to improve how AI coding agents interact with technical documentation. Unlike human users, AI agents fetch and process documentation through automated methods, often missing key engagement metrics traditionally tracked in analytics. AEO seeks to restructure and optimize documentation specifically for these agents, ensuring content is discoverable, parsable, token-efficient, and effectively indicates API capabilities. Failing to adapt documentation for AI agents can lead to either incomplete task execution or incorrect outputs due to limitations in context size. For the AI/ML community, AEO is significant as it aligns technical documentation with the rising reliance on AI agents for software development and integration tasks. Key practices include ensuring proper access through robots.txt, utilizing llms.txt for documentation indexing, and implementing skill.md files to clearly delineate API functionalities. By addressing issues with discoverability, parsing, and token management, developers can enhance the efficacy of their documentation for AI agents, ultimately leading to better performance in automated coding tasks. This evolution reflects a shift in developer experience and underlines the strategic importance of optimizing for machine interactions alongside human users.