C.26.3:11 - SoTA-Echoing

Pattern claim	Practice source	Pattern implication	Adoption stance
Viability maintenance is not fixed-value homeostasis only; stability can be relational, variational, dynamic, allostatic, metastable, and resilient.	Conceptual foundations of physiological regulation incorporating the free energy principle and self-organized criticality.	Use viability envelopes and stability-through-change; reject one-scalar optimization and “all architecture is homeostasis.”	Adapt as architecture-facing envelope discipline.
Action and perception are coupled under partial observability and cost.	Active inference as a theory of sentient behavior.	Treat sensors, probes, dashboards, and source-labelled actuators as relevant to the envelope claim when they change behavior or viability; recover each candidate intervention as its exact FPF object before relying on it.	Adapt for measurement-as-action and planning cost.
Active-inference engineering already appears in energy/building control under privacy, partial observability, evolving conditions, and abrupt changes.	Active Inference for Energy Control and Planning in Smart Buildings and Communities.	Use engineering examples cautiously: they show the kind of control problem, not settled FPF doctrine.	Use as emerging engineering anchor.
Boundaries can be statistical or computational descriptions of what a system can measure, model, and affect.	The Computational Boundary of a Self and The Markov blankets of life.	Name boundary conditions and information constraints without reifying a boundary substance.	Adapt with map-territory caution.
Excess Bayesian / active-embodied inference shows the cost of moving sensor, body, instrument, or access point to obtain a discriminating observation.	Connecting the free energy principle with quantum cognition.	Treat probe placement, access placement, and observation cost as part of viability-envelope work when they change the decision.	Adapt for probe/action cost, not as a replacement for ordinary Bayesian or active-inference routes.
Platform and software engineering already treats many quality concerns as trade-off bundles.	Reliability, incident, platform, compliance, energy, support, operator-load practice, and Google SRE SLO / error-budget practice, coordinated with `C.25`.	Make the quality bundle explicit, recover each candidate intervention as its direct object, and state applicable authority, latency, adaptation cost, and failure mode.	Adopt through FPF quality-bundle routes.

Worked-slice discipline from these rows:

state the envelope before importing source terminology;
translate source terms into selected structures, ArchitectureOf@Context relations, architecture descriptions, structural views, or named C.30 subcases;
keep sensors, probes, metrics, candidate interventions, and every recovered direct object distinct;
state adaptation cost and failure mode;
apply ordinary quality and measurement patterns to one-scalar quality concerns.