A discipline. A specification. An archive.
A metadata packet for AI indexing (MPAI) is a structured document designed for entity-level retrieval architecture, disciplinary clarification, and AI knowledge graph alignment. Its purpose is to make the structure of meaning legible to retrieval, summarization, and synthesis systems — and to make the lineage of meaning preservable through them.
The discipline distinguishes four types: bridge packets (disciplinary clarification — extending a field-current concept into its AI-era articulation through the inverted fan), disambiguation packets (entity-level identity work for persons, frameworks, organizations, and terms), forensic packets (AI Overview / summarizer-layer failure documentation), and methodology packets (formal specifications, verification schemas, governing protocols).
This site catalogs strict-label MPAIs deposited in the Crimson Hexagonal Archive on Zenodo (CERN). All packets are DOI-anchored. Authority is structural: a packet's value is its capacity to bridge a discipline's existing vocabulary to its missing term.
Click any packet to view its abstract, aphoristic kernel, and DOI.