# AFR.IM — AI-Friendliness Rating

> AFR.IM publishes objective AI-friendliness ratings for websites: an open, reproducible 0–100 AFR score and an A–F grade showing how open each site is to AI crawlers and agents. Every point traces back to public evidence anyone can fetch again — robots.txt policy, llms.txt, sitemap, structured data, semantic HTML, public APIs, and agent instructions. English at https://afr.im/ and Chinese at https://afr.im/zh-CN/.

## Pages

- [Home](https://afr.im/): directory homepage — grade distribution, latest scans, and the natural leaderboard sorted by pure AFR score.
- [Method](https://afr.im/method): the full AFR standard — three layers, every sub-check and its points, grade thresholds, and the neutrality pledge.
- [Categories](https://afr.im/categories): browse rated sites by category.
- [Submit](https://afr.im/submit): request a scan of a site you care about.
- [About](https://afr.im/about): who runs AFR.IM and how to get in touch.
- [For bots](https://afr.im/bot): what our own scanner fetches and how a score is produced.
- [Badge](https://afr.im/badge): embeddable rating badge for your site.
- [Search](https://afr.im/search): search the directory.
- [Offline fallback](https://afr.im/offline): cached directory view shown when the network is unavailable.

## API — human-readable documentation

- [API documentation](https://afr.im/docs): every public endpoint, parameters, error shapes, and example JSON responses.
- [Machine-readable AFR standard](https://afr.im/api/v1/method): layers, per-check points, and grade thresholds as JSON.
- [OpenAPI specification](https://afr.im/openapi.json): OpenAPI 3.1 description of the API (mirror at /.well-known/openapi.yaml).

## API endpoints

- [GET /api/v1/sites](https://afr.im/api/v1/sites): paginated directory listing with category, grade, search, and sort filters.
- [GET /api/v1/stats](https://afr.im/api/v1/stats): totals, grade distribution, and average score.
- [GET /api/v1/categories](https://afr.im/api/v1/categories): category tree with per-category site counts.
- [GET /api/v1/leaderboard](https://afr.im/api/v1/leaderboard): top sites by AFR score descending (never mixes in sponsored placements).
- [GET /badge/:domain.svg](https://afr.im/badge/example.com.svg): rating badge image for a domain.

## Optional: agent usage policy

- [Agent instructions](https://afr.im/agents.md): how to identify yourself, fetch budget, what content may be reused, and how to cite a rating.
- [AI crawler policy](https://afr.im/ai.txt): short machine-readable summary of what AFR.IM allows and why.
- [robots.txt](https://afr.im/robots.txt): explicit Allow rules for the 14 AI user-agent families plus the Sitemap declaration.
- Ratings, grades, and check names may be reused with attribution to AFR.IM; the AFR standard text is CC BY 4.0. See [About](https://afr.im/about) for contact and correction channels.