I built a Chrome extension that shows what AI agents actually see on your site — and it made me rethink GA4 completely

I've been running an API-first site for a few months (AgentShare — DeFi data for AI agents).

GA4 shows almost nothing. But my server logs tell a completely different story: GPTBot, ClaudeBot, dozens of MCP clients — they're calling my API every single day.

GA4 is blind to AI agents. They don't load JavaScript, they don't click buttons, they don't fire analytics events. They just read JSON and call endpoints.

So I built a Chrome extension that shows what AI agents actually see when they visit any website:

Agent Readiness Score (0–100) — based on robots.txt, llms.txt, ai.txt, and MCP discovery

AI crawler policy table — shows what bots like GPTBot, ClaudeBot, and PerplexityBot are allowed to do

Prompt injection scanner — detects hidden text and invisible elements that might hijack AI agents (new in v0.5)

MCP Connect — copy-paste config for Cursor, Claude Desktop, VS Code, Windsurf

The extension runs entirely client-side. No content is uploaded to any server.

What I learned: a low score doesn't mean your site is "bad." It means AI agents might be ignoring your content because the structure isn't machine-friendly. If that matters to you, this might help.

Link: https://chromewebstore.google.com/detail/nimndnhajfkicbnipbfdkmgencjejjed

Would love feedback from other API-first builders:

What metrics do you think matter most for AI visibility?

How are you tracking traffic that GA4 can't see?

Building in public, one awkward step at a time. 🚀