Is your documentation ready for AI agents?
When an agent can't use your docs, developers don't get an error. They reach for a competitor. Find out where you stand across 
Cursor, 
Claude and 
ChatGPT.
Developed by
in partnership withDachary Carey.
At OpenRouter we see firsthand how much traffic comes through AI coding agents. If your docs aren't well-indexed and agent-optimized, you're invisible to the fastest-growing buyer segment.
Making our docs agent-ready was one of the best investments we made. When Cursor and Claude can read your API reference cleanly, developers ship integrations without ever opening a browser tab.
Agent score directory
How the top API documentation sites score on agent-readiness
Your docs have a new reader: AI agents
AI coding agents are reading your API docs millions of times a day. If your docs aren't optimized for these agents, you're invisible to the fastest-growing segment of your user base.
Agent Score is the first industry benchmark for AI-agent readiness. Think Lighthouse, but for how effectively AI agents can discover, parse, and use your documentation. 0 to 100, 22 checks, 7 categories.
API platforms with higher agent readability are already seeing outsized adoption. Agent readiness isn't a nice-to-have. It's the new SEO. 73% of top API docs still lack an llms.txt.
Built on the open-source Agent-Friendly Docs Spec with no gatekeeping, no black boxes. Every check is transparent and community-driven, by Dachary Carey.
Here's what humans are saying about it
My agent couldn't use the docs and I started digging into why. The result is AFDocs, a standard that codifies what agents need to help developers complete their tasks. Agent Score makes that standard measurable.
Making our docs agent-ready was one of the best investments we made. When Cursor and Claude can read your API reference cleanly, developers ship integrations without ever opening a browser tab.
At OpenRouter we see firsthand how much traffic comes through AI coding agents. If your docs aren't well-indexed and agent-optimized, you're invisible to the fastest-growing buyer segment.
Postman's agent mode, like all agents, relies on LLM-ready docs to use APIs accurately with less hallucination. Agent Score gives you a concrete target to close that gap.
Docs aren't read by humans anymore: they're consumed by agents. AX has replaced DX.
We spend a lot of time on our content, but it doesn't matter if agents can't discover it. Agent Score is a beautifully designed tool that tells us how we suck.
Like moving from punch cards to IDEs or machine code to syntactic languages, agents are the next big leap in software engineering. Agent experience is now at the forefront of how products are adopted and it all starts with good, agent-first documentation.
Incident response is about speed. Rootly invested early in being agent-ready, but had no way to measure how far ahead we actually were. Agent Score changed that.
Developers who discover us through AI agents convert at a higher rate than any other channel. That's why our agent-readable docs are a key growth lever.
22 checks across 7 categories
A comprehensive framework for evaluating agent-readiness
