AGDAU: Ontological Admissibility Constraints for Safe Autonomous and General Intelligence

This document fixes the canonical formulation of AGDAU at the ontological level. Core claim: AGDAU is a necessary, but not sufficient, condition for the safe existence of autonomous and general intelligence. The document provides: a canonical statement of the AGDAU principle, a fixed set of ontological admissibility invariants (I1–I7), an abstract enforcement structure compatible with certifiable, safe-by-construction intelligent systems. Files included: PDF (human-readable canonical document), TXT (plain-text canonical copy for archival and machine processing).