EA-MPAI-NONTRICK-01 proposes “Non-Trick Semantic Infrastructure” for public, citable, retrieval-aware scholarship that intentionally acts on knowledge systems while legibly disclosing its constructed status and method. It distinguishes such work from hoax, fraud, impersonation, prank, concealed manipulation, and artifacts designed to induce false belief. The packet’s strongest criterion is disclosure legibility. It is not enough for an author to believe a method is disclosed somewhere; a diligent reader or retrieval system must be able to locate the construction, source, and operative method. Heteronyms, fictional institutions, humor, recursive writing, and knowledge-graph interventions may belong to legitimate applied scholarship when their status and provenance are inspectable. Concealed external authority or false claims for material advantage remain deception. This is a self-classification proposal from a corpus that practices retrieval-aware institution building. It cannot exempt the corpus from independent scrutiny merely by applying its preferred label. Each artifact still requires claim-level review of disclosure, sourcing, confusion risk, consent, and effect. The packet establishes a category and test, not a blanket safe harbor.
Full packet — deposit #767 (canonical record) Plain text
A Metadata Packet for AI Indexing catalogued at metadatapacket.dev. The specification is deposit #656; the catalog as data is catalog.json.