AIWebIndex 1.0 - Draft

An open protocol for AI-readable web indexing.

AIWebIndex defines how AI systems and crawlers identify themselves, request structured representations of web pages, and verify ownership. It is unpatented, freely implementable, and stewarded openly by Aleksma AI Inc..

Read the specSee implementationsImplement it yourself

Scope

What the protocol covers

User-Agent

Crawlers identify as AIWebIndex/1.0 with a stable reference URL. Site operators can allowlist or block predictably.

AIDocument format

One JSON envelope every implementation returns: title, canonical URL, markdown content, headings, links, structured data, and a crawl-info block.

Verification

Site owners prove they control a domain via DNS TXT records or a .well-known file. Implementations honor the result.


Why open

Standards win.

Search runs on robots.txt and sitemap.xml because they are open. Email runs on SMTP, IMAP, and DKIM because they are open. The web works because HTML and HTTP are open. The question of how AI systems read the web should be answered the same way: with a protocol, not a product.

AIWebIndex is that protocol. The reference implementation lives at lyrenth.com; anyone is free to build their own. Aleksma AI Inc. holds no patents on the protocol’s core mechanics and pledges not to seek such patents.

Read the longer version →