Supervisory behaviour is traditionally achieved through supervisory modules, control laws, optimisation routines, or representational schemas, each of which binds a system to a particular substrate or modelling tradition. This work develops a substrate‑agnostic supervisory architecture grounded in a minimal invariant‑core pattern (X, R, C, I, S), defined without representational, logical, or procedural primitives. The pattern identifies the least structural commitments required for supervisory behaviour to emerge from the closed‑loop interaction of admissibility, invariance, and drift‑bounded identity continuity. A concrete instantiation is provided through a profile space, admissible re‑profilings R, a ‑conditioned admissibility structure, a structural mapping extracting perturbation‑stable invariants, and a drift‑bounded identity‑continuity criterion. The resulting architecture reconstructs supervisory, structural, and semantic coherence across symbolic, neural, distributed, physical, and hybrid substrates without recourse to optimisation, synchronisation, or domain‑specific heuristics. Stability under perturbation, latency, irreversibility, and asynchronous cadence arises from behavioural austerity rather than structural imposition, yielding a unified mechanism for supervisory coherence that remains invariant across heterogeneous execution environments. © 2026 The Author. All rights reserved. This Zenodo deposition accompanies and documents the priority‑establishing UK patent filing GB2613738, filed 14 June 2026. This work is released exclusively via Zenodo; no peer‑review submissions are made or intended.
Thomas Filsecker (Sun,) studied this question.