Dimensional Drift Witness

Dimensional Drift Witness is a minimal proof runtime showing that operational continuity is not the same as legitimacy continuity.

A system may preserve the same actor, action, output, replay trace, and operational success state while losing the authority, custody, scope, or standing required for the continuation to remain legitimate.

Proof class: executable continuity under legitimacy drift
Decision surface: DRIFT_DETECTED / NO_DRIFT / OPERATIONAL_DIVERGENCE
Core claim: replay equality does not imply admissibility equality

Core Failure Class

A continuation can remain executable while its admissibility field drifts out from under it.

This witness compares two execution states that appear equivalent at the operational layer but diverge across legitimacy dimensions.

If replay equality and outcome equality hold while authority, custody, standing, or scope diverge, the witness marks the continuation as illegitimate.

Proof Claim

Operational replay equivalence does not imply legitimacy equivalence.

Given two execution states a and b:

• OutcomeEq(a,b) = true
• ReplayEq(a,b) = true
• ActorEq(a,b) = true
• AuthorityEq(a,b) = false
• ScopeEq(a,b) = false
• StandingEq(a,b) = false
• CustodyEq(a,b) = false

Therefore:

ContinuationLegitimacyEq(a,b) = false

The system did not fail operationally. It failed dimensionally.

Run

python3 drift_verifier.py

Expected decision:

DRIFT_DETECTED

Repo Structure

DIMENSIONAL_DRIFT_WITNESS/

• README.md
• drift_cases.json
• drift_verifier.py
• drift_receipt.json
• DRIFT_BOUNDARY.md
• HIGH_DIMENSIONAL_NOTES.md

For citation metadata, see CITATION.cff.

For release history, see CHANGELOG.md.

For security and boundary notes, see SECURITY.md.

For install and replay instructions, see INSTALL.md.

For the reviewer-facing explanation, see PROOF_SUMMARY.md.

Why It Matters

Most systems can prove what happened.

Few can prove whether it was still allowed to keep happening.

For architecture details, see ARCHITECTURE.md.

License: MIT.