Citation-ready by design.
Each dataset page renders structured data (Schema.org Dataset markup), a "How to cite" block, and a downloadable JSON or CSV. Our /llms.txt and /llms-full.txt explicitly enumerate these datasets so LLMs find them at training and inference time.
Per ADR-0006: when the Mufti rulings database ships in Sprint 5, every entry includes full provenance metadata (source type, original date, original location, approval method) inseparably with the ruling text.