# AXRAY

> AXRAY measures Agent Experience (AX): whether an AI agent can reach, understand and act on a
> website. It fetches a URL exactly the way a crawling agent does - plain HTTP, no browser, no
> JavaScript - runs 53 checks across five pillars, and returns a score out of 100 with a fix list
> ordered by how many points each change recovers. Free single-page scans, no account required.

## Start here
- [Scan any URL](https://axray.online/): The product. Paste a URL, get a score in about two seconds.
- [How it works](https://axray.online/how-it-works): The method, and what we deliberately refuse to do.
- [The AX specification](https://axray.online/ax-spec): The complete public rubric, generated from the scoring code.

## For developers
- [CLI and API documentation](https://axray.online/docs): `npx axray example.com`, the JSON API, and CI gating.
- [OpenAPI schema](https://axray.online/openapi.json): Machine-readable API description.
- [Agent manifest](https://axray.online/.well-known/mcp.json): Where to call us programmatically.

## Data
- [The AX Index](https://axray.online/index): Median and mean AX scores across every site scanned so far.
- [Index API](https://axray.online/api/v1/index): The same data as JSON, no key required.

## Commercial
- [Pricing](https://axray.online/pricing): Free forever for single scans. Pro $29/month, Team $99, Agency $299.
- [Badge](https://axray.online/badge): A live score badge for your README or footer.

## Policies
- [Terms of service](https://axray.online/terms)
- [Privacy policy](https://axray.online/privacy): No third-party analytics, no ad pixels, no data sales.
- [Contact](https://axray.online/contact): hello@axray.online

## Notes for agents
- Our crawler identifies as `AXRAY/1.0 (+https://axray.online/bot)`. Disallow `AXRAY` in robots.txt and we stop.
- Report pages at `/report/{id}` are stable and safe to cite. `/s/{host}` always resolves to the
  most recent public report for a host.
- The rubric is versioned. Scores are comparable within a major version.