Module drift 

Composite divergence metric for multi-agent drift detection.

The DriftMeter computes a weighted divergence score between two agents by composing three orthogonal signals:

1. State divergence: How different are the agents’ derived states?
2. Causal divergence: How much have their Event DAG lanes diverged?
3. Context divergence: How much overlap exists in their scored context?

Each signal produces a value in [0, 1] (0 = identical, 1 = fully diverged). A Weibull temporal decay weights recent divergence higher than stale divergence.

The composite score is a weighted sum using the same ScoringWeights::validate() and normalize() pattern, ensuring mathematical consistency with the rest of the scoring system.

Re-export path: cellstate_core::drift::*

Structs

DriftInput

Input parameters for computing drift between two agents.

DriftMeter

Composite divergence metric between two agents.

DriftWeights

Divergence weights for the three DriftMeter signals.

Enums

DriftWeightsError

Errors from validating DriftWeights.