Mechanism
Confidence coupling is the cross-primitive integration by which the integrity field serves as a direct input to the confidence-governed execution mechanism. The integrity field is a deterministic, multi-domain structure that encodes the agent's modeled alignment between its declared values and its actual behavioral record as preserved in lineage. The three integrity domains, personal, interpersonal, and global, are tracked independently and combined into a composite integrity score using domain weights specified by the applicable policy configuration. That composite integrity score is one of the input variables consumed by the confidence computation. The coupling is structural: integrity does not merely affect the agent's internal coherence state, it produces observable, consequential changes to the agent's execution authorization and speculative planning.
The disclosed consequence is concrete. When the agent's integrity score is degraded, due to recent deviation events, active coping intercepts, or accumulated restoration gaps, the confidence computation produces a lower confidence value, reflecting the reduced reliability of the agent's decision-making under compromised integrity conditions. The coupling therefore makes integrity reliability legible to the part of the architecture that decides whether the agent is permitted to act.
The Execution Mode Transition
If the integrity-modulated confidence value falls below the execution threshold, the agent transitions from executing mode to non-executing cognitive mode. In non-executing cognitive mode the agent does not commit actions, but it continues to forecast, construct planning graphs, and generate inquiry requests. This is the central operational effect of the coupling: an agent that deviates sufficiently loses the authorization to act and must instead pause, reflect, gather information, and plan until its integrity is sufficiently restored to support confident execution.
The restriction is on committed execution, not on cognition. The agent operating under compromised integrity continues cognitive processing but suspends committed execution until the internal conditions for reliable action are reestablished. Integrity is restored through redemption engine outputs or through environmental changes, and as it recovers the confidence value rises back above the execution threshold and execution authority returns.
Integrity-Modulated Forecasting
The same integrity state that conditions confidence also conditions the forecasting engine. The forecasting engine receives the agent's current integrity state as a conditioning input when constructing planning graphs. Under compromised integrity the engine is conditioned to weight conservative branches more heavily than aggressive branches, reflecting the reduced risk tolerance appropriate under compromised integrity; to extend the evaluation horizon for speculative branches, reflecting the need for more thorough assessment when the agent's judgment is compromised; to generate branches that include explicit integrity restoration steps, so recovery becomes part of the agent's forward-looking strategy; and to increase the pruning threshold for branches that involve high-integrity-risk actions.
Because committed execution is suspended while cognition continues, the agent in non-executing mode is precisely the agent that needs forecasting most. The coupling routes a low-integrity agent into a planning posture that is both more cautious and explicitly oriented toward restoring the conditions for action.
The Recovery Cycle
The integration between integrity, confidence, and forecasting produces a coherent behavioral cycle. Integrity violations trigger confidence reduction; confidence reduction triggers execution pause; execution pause triggers forecasting-based recovery planning; and recovery planning generates restorative mutations that, when eventually executed after confidence recovery, restore integrity. The cycle operates without external intervention: the agent's own cognitive architecture detects the integrity problem, restricts action, generates a recovery plan, and implements it when conditions permit.
This is a closed loop. The disclosure describes a feedback path in which the recovery cycle feeds back to the integrity field, so the agent that has restored integrity through executed restorative mutations recovers the confidence required to resume normal committed execution.
Where the Coupled Integrity Comes From
The integrity score that drives this coupling is not an external label. The integrity field is self-referential: the agent maintains its own integrity model based on its own actions, its own declared values, and its own policy constraints. Degradation in that score arises from the same subsystem that records deviation as truth. When the deviation function output exceeds the policy-defined activation threshold, the agent enters a Deviation-Activated State, deviation-class mutations are recorded in lineage with full provenance, and the integrity field is updated immediately. Coping intercepts, in which the coherence loop is interrupted at the empathy, integrity-recording, or self-esteem-restoration phase, and accumulated restoration gaps, the residual integrity loss the redemption engine could not address, are the other named contributors to a degraded score.
Because the integrity score is computed from the auditable lineage and feeds confidence deterministically, the loss of execution authority is itself traceable. An agent cannot selectively omit integrity events or present an integrity state inconsistent with its lineage without producing a detectable trust slope discontinuity, so the integrity value that lowered confidence can be reconstructed and verified against the recorded behavioral history.
Application to Discovery Traversal
The same coupling is applied to semantic discovery traversal. During traversal of the adaptive index, the discovery object maintains a traversal integrity field that records the degree to which the traversal remains aligned with the original query intent, and the integrity engine computes a semantic drift metric at each step. When a traversal integrity violation is recorded, it produces the same structural consequences within the discovery context that behavioral integrity violations produce in the agent context: confidence degradation that may trigger traversal pause and re-evaluation, forecasting activation that generates alternative traversal strategies, and integrity logging that creates an auditable record of the semantic drift.
The disclosure is explicit that this reuse is not an analogy but the same mechanism: the same deviation detection, drift measurement, and corrective pressure mechanisms that operate on agent behavior operate on traversal behavior because both are expressed as sequences of semantic mutations within the same governance framework. A traversal whose integrity has degraded may be redirected back toward the original intent, branched to explore the drift target as a secondary objective, or terminated with a confidence-qualified result that discloses the integrity violation to the consuming entity.
What the Coupling Is and Is Not
The coupling is not a confidence estimator and not an integrity estimator. It is the structural integration between the composite integrity score and the confidence computation, together with the operational consequence that follows: an execution mode transition gated on whether integrity-modulated confidence clears the execution threshold. Integrity defines reliability; confidence translates reliability into execution authority; forecasting fills the suspended interval with recovery planning.
The disclosure does not describe a graded envelope of permitted actions, a tunable transfer curve, or named authority tiers. It describes a binary execution mode boundary: above the execution threshold the agent commits actions, below it the agent suspends committed execution while continuing to forecast, plan, and inquire. The consequence of integrity loss is loss of the authorization to act, not a continuous reshaping of an action space.
Disclosure Scope
The cross-primitive integration of the integrity field with the confidence-governed execution mechanism and the forecasting engine, comprising integrity-modulated confidence in which a degraded composite integrity score lowers the computed confidence value, the transition from executing mode to non-executing cognitive mode when integrity-modulated confidence falls below the execution threshold, integrity-modulated forecasting that conditions planning-graph branch weighting and pruning under compromised integrity, the closed recovery cycle from integrity violation through execution pause and forecasting-based recovery planning to restorative mutation and integrity restoration, and the application of the same coupling to discovery traversal through the traversal integrity field and semantic drift metric, is disclosed in the cognition filing (U.S. Application No. 19/647,395 and its international counterpart). This article describes that disclosed mechanism. The scope does not extend to confidence estimation as such or to integrity computation as such; it covers the structural coupling between them and the execution-authority consequences that follow from treating the integrity score as a conditioning input to confidence and forecasting.
