Summary
Edmund Yong walks through a concise, three-part framework for making software applications discoverable and usable by AI agents — a growing concern as AI-powered search tools like ChatGPT Search and Perplexity increasingly mediate how users find and recommend products. The three areas covered are discoverability, legibility, and actionability.
On discoverability, Yong recommends researching competitor landing page language across Google, ChatGPT Search, and Perplexity, then updating headers and body copy to clearly define the app’s niche, target customer, integrations, and outcomes. Technical SEO basics are also covered: ensuring pages return 200 status codes, verifying robots.txt isn’t blocking major AI crawlers, and confirming sitemaps include key public pages. On legibility, the focus is semantic HTML — using proper button, link, form, and heading tags so agents can understand page structure without guessing — plus writing clear documentation for common workflows and error cases. On actionability, Yong covers exposing REST API endpoints with API key authentication, input sanitization, and rate limiting, and then building MCP servers and CLIs for developer-facing tools.
The video also features CodeRabbit, an AI code review tool whose new “cohorts” feature splits large pull requests — a common byproduct of AI-assisted coding — into logically grouped chunks for easier review, alongside inline code inspection via Code Peek and a follow-up question interface called Agentic Chat.
📺 Source: Edmund Yong · Published July 06, 2026
🏷️ Format: Tutorial Demo
