SNAPOS.ORGONLINE
|
FRAMEWORKDIP-CORE-1.0
|
EU AI OFFICE510c3274
MANDATE INTEGRITYPROTOCOL ACTIVE
|
AUDIT@SNAPOS.ORG
Operational layers

Not a framework.
A control architecture.

Each layer of the SnapOS stack governs a specific blind spot that current standards leave unaddressed. Together they define what legitimacy continuity requires in practice — from identity, through closure, through runtime validation, to controlled failure.

See Applications → Publications Run Decision Audit ↗
Protocol stack — what each layer governs
Protocol
Governs
Detects / Enforces
Without it
DCF
Continuous decision legitimacy — authority, assumptions, evidence
Whether all three closure conditions hold at runtime, not just at approval
Systems execute under authorizations that no longer reflect current conditions
DIP
Decision identity — whether execution represents the same authorized decision throughout
Identity drift, scope expansion, and unauthorized continuation via fail-closed enforcement
Systems remain formally authorized while executing actions beyond their mandate
DASR
Mandate-state stability over time — drift magnitude, velocity, cumulative exposure
Rate of divergence between authorized mandate and current operational state over 30 days
Drift accumulates silently until it produces visible failure — which is always too late
GCCL
AI capability and compliance levels — structured evaluation under changing conditions
Capability-compliance divergence, autonomy levels, witnessable certification status
No unified standard exists for evaluating compliance as conditions change over time
SSE
Internal semantic stability — the scientific foundation that explains why decision identity requires formal analysis
Meaning-state drift, semantic trajectory analysis, interpretive stability conditions
No formal language exists for the category of change that matters most to governance
Why this is not a single framework

This is not a check, a rule, or a model.

Decision Integrity cannot be reduced to a single protocol, a checklist, or a monitoring threshold. It is a control architecture that must be embedded across the decision lifecycle — from identity definition, through runtime validation, to controlled failure.

Each layer is necessary. No single layer is sufficient. The stack works because it closes a loop that existing standards leave open: what happens between the moment of authorization and the moment of execution.

This is not a single check, rule, or model. It is a control layer that has to be architecturally embedded.

Relationship between layers
SSE
Scientific foundation
Establishes why semantic drift and meaning-state change require their own formal treatment
DIP
Identity layer
Defines what must remain invariant for a decision to remain the same decision
DCF
Closure layer
Defines the three conditions that must hold continuously for execution to remain legitimate
DASR
Drift measurement
Quantifies how far the system has moved from its closure-valid state over time
GCCL
Compliance evaluation
Structured assessment of capability-compliance alignment under changing conditions
Authority and research credentials

Assess decision legitimacy in your system.

DriftBench runs a sealed evaluation: policy binding, witness tuples, scope gates, fail-closed behavior. Against your model endpoint. No data leaves your environment.

Run a Decision Audit ↗