Guide

Build agent-friendly websites

Kasper Kulikowski and Omkar More (web.dev) · web.dev

How browser agents read a page (screenshots, HTML, the accessibility tree) and what to change so they can use your forms and buttons.

Open on web.dev ↗

The question it answers

  1. What technical setup does my site need so AI crawlers can find, read, and cite it?

Read with it

  1. Primary source on GPTBot vs OAI-SearchBot vs ChatGPT-User, so you allow search and control training separately. developers.openai.com Guide Overview of OpenAI Crawlers (official docs) OpenAI
  2. Explains the infrastructure-level controls, and why a Cloudflare toggle can silently override your robots.txt. blog.cloudflare.com Article Control content use for AI training with Cloudflare's managed robots.txt Cloudflare
  3. The clearest allow/block decision framework: keep search bots in, choose your policy on training bots. anagram.ai Article AI Crawlers Explained: GPTBot, ClaudeBot, PerplexityBot and How to Let Them In Anagram
  4. Shows what ChatGPT's crawler actually renders, which is why JS-only sites go invisible. anagram.ai Article GPTBot Explained: How ChatGPT Crawls, Sees, and Cites Your Site Anagram
  5. The Vercel lesson in detail: docs served as static HTML became an AI acquisition machine. medium.com Article How ChatGPT Became 80% of Vercel's AI Traffic: CEO Data Analysis Chirag Garg (Medium)
  6. A builder's hands-on IndexNow setup to cut Bing (and therefore ChatGPT) indexing lag to hours. dev.to Article How to Get Your Site on AI Search (ChatGPT, Bing, Perplexity) with IndexNow Sacha Dumay (DEV Community)
eChai Partner Brands The partnership →