Annex C — Deterministic Replay Equivalence

Trust-State Standard v1.0
Status: Official Release
Publication Date: February 2026
Published by VTI Foundation, Inc.

This annex defines deterministic replay equivalence requirements for Trust-State evaluations.

1. Scope

Deterministic replay ensures that identical evidence and rule identity produce identical evaluation outcomes across independent executions.

2. Replay Inputs

Replay SHALL be executed using:

Canonical evidence byte sequence
Derived RULE_HASH
Invariant evaluation logic bound to the rule definition

3. Replay Equivalence Condition

Replay equivalence SHALL evaluate TRUE if and only if:

(recomputed_evidence_hash == original_evidence_hash)
AND
(recomputed_rule_hash == original_rule_hash)
AND
(recomputed_decision == original_decision)
AND
(recomputed_state == original_state)

4. Determinism Requirement

Evaluation logic SHALL be free of:

Non-deterministic randomness
External time dependencies
Uncontrolled environmental variables

Any non-deterministic influence SHALL invalidate replay equivalence.

5. Failure Condition

If replay equivalence evaluates FALSE under identical canonical inputs and rule identity, the implementation SHALL be considered non-conformant.

6. Relationship to Certification

Deterministic replay equivalence is a mandatory condition for Trust-State conformance and certification eligibility.