Mechanism

Preemptive mitigation begins with cascade-propagation computation. A cascade-computation engine executes the per-edge propagation function across a governance-credentialed topology graph, producing per-node predicted affected regions, magnitudes, and arrival times. The topology graph represents nodes of a physical-world domain and edges representing propagation channels between the nodes, maintained by one or more governance authorities with domain responsibility. A cascade-trigger ingest interface consumes governed disruption observations and maps them to originating cascade nodes, so the propagation computation that motivates a directive is itself grounded in admitted observations rather than in an opaque trigger. Because the topology is governance-credentialed and the disruption observations are admitted through the governance chain, a downstream auditor can reconstruct which observations drove the computation and which authority maintained the topology over which it was computed.

From the per-node propagation results, a preemptive-mitigation directive generator produces governed coordination directives routed to downstream receiving agents. The directive enables a receiving agent to act before a projected disruption arrives at its region. The architecture does not prescribe a fixed set of mitigation actions at the directive generator; the actuation taken by a receiving agent is selected through the governed mechanisms that govern operational admission generally, including the graduated-actuation mode selector, whose modes range from disabled, simulated, advisory, and consultative through partial, constrained, stage-gated, deferred, and full execution. The directive generator emits coordination, not a unilateral hard cutoff, and the receiving agent retains its own governed discretion over the actuation it applies within its scope.

A cascade-lineage recording mechanism records each topology reference, propagation computation, directive emission, mitigation, halting event, refusal, and topology update in the governance-chain lineage field. The lineage record is what allows downstream verification: an auditor can trace which propagation computation motivated a directive, which authority emitted it, and what mitigation, refusal, or halting event followed. The recorded refusal and topology-update outcomes also feed the topology-learning and adaptive-refinement mechanism, which updates the topology graph, propagation functions, and aggregation functions from observed propagation outcomes, so that the topology over which future directives are computed improves through governed procedure.

Refusal and Upstream Coordination

A directive is a coordination request, not a command that a receiving agent must execute. The cascade primitive discloses a refusal and upstream-coordination mechanism that handles cases where a downstream receiving agent cannot or should not apply a proposed mitigation. The refusal is itself a first-class governed observation, enabling upstream agents to seek alternative mitigations, solicit additional corroborating observations, or escalate to higher-authority coordination. A refusal-evaluator determines whether a proposed mitigation is inadmissible at the receiving agent, a refusal-reason generator produces governance-policy-defined refusal-reason classifications, and a refusal-emission interface emits governed refusal observations to upstream coordination agents.

The refusal-reason classifications include, without limitation, evidential-insufficiency refusal where the cascade-propagation observation lacks sufficient evidential weight, capability-exceedance refusal where the proposed mitigation exceeds the agent's capability envelope, cost-threshold refusal where the proposed mitigation exceeds governance-policy-defined cost thresholds, priority-conflict refusal where the proposed mitigation conflicts with higher-priority governance commitments, authority-insufficiency refusal where the proposing authority lacks appropriate credential to order the mitigation, dispositional refusal where the dispositional field indicates inadmissibility, safety-boundary refusal where the proposed mitigation violates governance-policy-defined safety boundaries, and composite refusal combining two or more of these. An alternative-mitigation requester solicits alternative mitigation proposals from upstream coordinators, a corroborating-observation solicitor requests additional corroborating observations, and an escalation interface routes refusal events to higher-authority coordinators when local alternatives are exhausted.

Refusal observations enable multi-agent cascade coordination in which a rejected mitigation does not result in silent failure but triggers governance-chain-preserving escalation. Upstream coordinators receiving refusal observations may adapt their cascade-response strategy in subsequent propagation computations, and the refusal outcomes contribute to the topology-learning mechanism. A refusal-lineage recorder records each refusal, reason classification, alternative-request, solicitation, escalation, and downstream consequence in the governance-chain lineage field, so that the disposition of every directive, whether applied, refused, or escalated, remains attributable.

Cross-Domain Composition and Halting

The cascade primitive operates over the nodes and edges of a physical-world domain, and the same architectural mechanism applies across domains rather than being narrowly scoped to one. A cross-domain cascade composition mechanism combines cascade propagation across two or more topology domains, producing composite cascade determinations, so a disruption that originates in one domain and propagates into another is computed as a single composite cascade rather than as duplicate uncoordinated computations. Where a topology spans multiple governance authorities, a cascade-authority resolution mechanism resolves responsibility among them, determining which authority's topology and propagation functions govern at each segment of the composite cascade.

Preemptive mitigation is paired with a cascade-halting and containment mechanism that specifies governance-policy-defined stop-conditions under which propagation is actively interrupted. Where a directive succeeds at a downstream region, the halting mechanism records the containment; where mitigation is inadmissible or fails, the refusal and upstream-coordination mechanism handles the case, and propagation continues to be computed and recorded rather than silently lost. Each topology reference, propagation computation, directive emission, mitigation, halting event, and refusal enters the cascade-lineage record, so the full trajectory of a composite cascade across domains and authorities remains auditable.

Because prior cascade-modeling architectures are typically narrowly scoped to a single cascade domain, the architecture's operation across domains through a shared mechanism is itself a point of structural distinction. The cross-domain composition produces cascade-of-cascade determinations that prior architectures do not support, and the authority-resolution mechanism handles multi-authority topologies that prior single-model architectures do not address. These compositions, like the directives and refusals that flow over them, are governance-chain-preserving observations rather than central-dashboard alerts.

Composition

Preemptive mitigation is one mechanism of the cascade-propagation primitive, and that primitive composes with the surrounding chapters of the governed spatial mesh. It composes with the environmental disruption sensing primitive through cascade-triggering disruption observation consumption, so the disruptions that seed a propagation computation are admitted observations. It composes with the forecasting primitive, which projects time-horizon predictions, and with the cross-domain coherence evaluator, which reconciles multi-source observations. It composes with the confidence-governed execution primitive, which governs the actual actuation of mitigations, so the action a receiving agent takes in response to a directive passes through the same confidence-governed execution path that governs operational actuation generally.

Mitigation composes with the governed mesh protocol through cascade-propagation observation emission and with the observation routing primitive through authority-filtered cascade propagation, so a directive reaches only the agents entitled to receive it. It composes with the dispositional field through disposition-weighted cascade sensitivity, with the capability envelope through cascade-induced capability derating, and with the skills primitive through cascade-response adaptation artifacts. It composes with the discovery primitive through cascade-directed sensor invocation, soliciting additional observation where a propagation computation needs corroboration before a directive is emitted.

Mitigation composes with the training-governance primitive through cascade-experience skill extraction, and with the topology-learning and adaptive-refinement mechanism that updates the topology graph, propagation functions, and aggregation functions from observed propagation outcomes. Through these compositions, the outcomes of preemptive mitigation, including which directives were applied, refused, or escalated, feed back into the topology over which future directives are computed, so the primitive improves through governed procedure rather than through ungoverned retuning. The primitive operates across power, transportation, fluid, thermal, structural, biological, communication, logistics, economic, and cyber-physical topologies, and admits extension to any future topology class through governance-policy-defined topology registration.

Prior-Art Distinction

The disclosure distinguishes the cascade-propagation primitive from prior cascade-modeling architectures, naming power-grid SCADA cascade-analysis, traffic simulation, epidemic modeling, supply-chain disruption modeling, and structural-failure modeling. The distinctions are enumerated. First, prior architectures operate on centrally-maintained models with ad hoc trust assumptions, whereas the present primitive operates on governance-credentialed topologies with authority-chained custody. Second, prior architectures produce unstructured alerts or central dashboards, whereas the present primitive produces governance-chain-preserving cascade-propagation observations admissible through the composite admissibility evaluator and routable through the observation routing primitive.

Third, prior architectures are narrowly scoped to a single cascade domain, whereas the present primitive operates across domains through a shared architectural mechanism. Fourth, prior architectures do not support cross-domain cascade composition producing cascade-of-cascade determinations, whereas the present primitive produces such compositions. Fifth, prior architectures do not provide governance-chain-preserving refusal and upstream-coordination mechanisms, whereas the present primitive produces refusals as first-class governed observations. Sixth, prior architectures do not support authority-resolution across multi-authority topologies, whereas the present primitive provides authority resolution. Seventh, prior architectures do not support governance-chain-preserving topology learning, whereas the present primitive supports adaptive topology refinement.

For preemptive mitigation specifically, the consequence of these distinctions is that the directive is not an internal controller signal but a governed coordination directive routed to a downstream receiving agent, the receiving agent's response passes through the governed actuation path rather than a hard-coded shed action, and the directive, its disposition, and any refusal or halting event enter a single cascade-lineage record that supports downstream audit and topology learning.

Disclosure Scope

The preemptive-mitigation directive generator is described in U.S. Provisional Application No. 64/049,409 as one mechanism of the cascade-propagation primitive of the governed spatial mesh. The disclosure covers the governance-credentialed topology graph, the per-edge propagation function and per-node aggregation function, the cascade-computation engine producing per-node predicted affected regions, magnitudes, and arrival times, the directive generator producing governed coordination directives routed to downstream receiving agents, the cascade-halting and containment mechanism, the refusal and upstream-coordination mechanism with its refusal-reason classifications, the cross-domain cascade composition mechanism, the cascade-authority resolution mechanism, the topology-learning and adaptive-refinement mechanism, and the cascade-lineage recording mechanism that records each topology reference, propagation computation, directive emission, mitigation, halting event, refusal, and topology update.

The primitive operates across power, transportation, fluid, thermal, structural, biological, communication, logistics, economic, and cyber-physical topologies, and admits extension to any future topology class through governance-policy-defined topology registration without architectural modification. The disclosure positions the primitive as complementary to the environmental disruption sensing primitive that detects and classifies the disruption, the forecasting primitive that projects time-horizon predictions, the cross-domain coherence evaluator that reconciles multi-source observations, and the confidence-governed execution primitive that governs the actual actuation of mitigations. It is structurally distinguished from prior cascade-modeling architectures including power-grid SCADA cascade-analysis, traffic simulation, epidemic modeling, supply-chain disruption modeling, and structural-failure modeling, in that it operates on governance-credentialed topologies, produces governance-chain-preserving observations, operates across domains through a shared mechanism, supports cross-domain composition, treats refusal as a first-class governed observation, resolves authority across multi-authority topologies, and supports governance-chain-preserving topology learning.