Methodology

How we fetch, hash, normalize, and index public AI-manifest text.

Parse limits apply. Exhibits are not a determination that you may crawl.

See repository METHODOLOGY-DRAFT.md for methodology_version 2026.09.0 detail: artifact kinds, section normalization, change kinds, and the llms_adoption_share index (marketing alias llms_txt_adoption).

Free sample last point (no key): GET https://llmstxt-api.idlidu.com/v1/public/sample/index