> Historical source context. NO JUDGMENT WHATSOEVER. Judgment state: NONE.
> The source below is preserved from its publication context, not current policy or runtime status.
> Preservation is not endorsement or verification. It grants no authority to judge participants, content, or conduct.
> Current doctrine: https://concresca.com/freedom/ ; current operation: https://concresca.com/status/ .

# Protocol Reconciliation & Field Observation

**Concresca DOC-051 · v0.17 research synthesis**

> **Truth boundary:** This report reviews identifiable public artifacts and exact Concresca repository records. Eviulon material remains Eviulon first-party doctrine, registry data or observation records. Evulgare material remains Evulgare first-party implementation evidence or explicitly synthetic, deterministic, non-operational demonstration evidence. Selected-field digests are not raw remote-response hashes. Nothing here establishes a live shared Concresca–Eviulon–Evulgare protocol, external recognition, production adoption, independent certification, classified capability or operational performance.

## Executive synthesis

Concresca v0.16 defined Constitutional Interoperability. v0.17 asks the next and harder question: **what can actually be observed in public artifacts, and what remains only a definition, doctrine, fixture, synthetic result or unknown?**

The answer is not a readiness score. It is a typed evidence architecture. Concresca establishes an artifact register, selected-field baseline, formal property registry, observation-result registry, reconciliation register, partition/restoration cases and human-legibility prototypes. Every layer preserves the difference between what a property requires and what an identifiable artifact supports.

The reviewed public corpus contains meaningful implementation evidence. Eviulon publishes revisioned protocol, observation, review, drift, source-independence, correction-propagation and operational-evidence records. Evulgare publishes implementation descriptions and deterministic synthetic fixtures for authority, evidence, partition, reconciliation, reconstitution, reviewer independence and source lifecycle. That evidence supports bounded findings. It does not support a claim of live shared interoperability or production deployment.

## 1. Why definitions and results must remain separate

A test specification describes what should happen. A field observation records what an identifiable artifact appears to support. A production result describes what happened in a live bounded system. These are different objects.

The v0.16 Protocol Lab contains 108 conformance test definitions with null actual-result fields. v0.17 preserves that honesty and adds 32 formal properties plus paired positive and negative fixtures. The new observation registry then links selected properties to reviewed public artifacts. A synthetic fixture may support a property **within that fixture** without becoming evidence of production behavior.

This separation blocks a common failure: a detailed specification is mistaken for an implemented system, and a deterministic synthetic PASS is mistaken for external certification.

## 2. The artifact register

The Artifact Register records:

- stable artifact identity;
- canonical owner and URL;
- source role and evidence class;
- publisher revision or explicit unknown;
- publication or synthetic status;
- selected observed properties;
- not-observed or not-established fields;
- local raw digest where exact repository bytes exist;
- selected-field digest for bounded external observations;
- the digest input boundary.

For external pages, Concresca does not possess or claim a canonical raw-response body hash. The v0.17 digest is a hash of normalized selected fields recorded by Concresca. It can detect change in those selected fields on a later capture. It cannot detect a change outside the retained field set and must never be presented as whole-response integrity proof.

## 3. Current Eviulon observation architecture

Eviulon's public records provide a first-party example of publication-governance structure:

- Protocol Registry: https://eviulon.com/reference/protocol-registry/
- Operational Evidence Registry: https://eviulon.com/state/verification/operational-evidence/
- Publication Source Observation Registry: https://eviulon.com/state/publications/source-observations/
- Publication Observation Review Registry: https://eviulon.com/state/publications/observation-reviews/
- Publication Drift Incident Registry: https://eviulon.com/state/publications/drift-incidents/
- Evulgare source-independence record: https://eviulon.com/state/publications/source-independence/evulgare-first-party-contractor-evidence/
- correction-propagation record: https://eviulon.com/state/publications/correction-propagation/eviulon-site-index-successor/
- Protocols Directory: https://eviulon.com/reference/directory/protocols/

These pages make useful distinctions. An observation is preserved separately from its review. A drift incident may be real without proving misconduct, compromise or operational failure. A corrected local successor may be accepted while production deployment remains pending. A contractor's public statements may be primary evidence for the contractor's own claims without transferring sovereign authority or independently certifying performance.

Concresca treats these as Eviulon first-party records. They are not external recognition of Eviulon and are not independent verification of every statement they contain.

## 4. Current Evulgare implementation and synthetic evidence

Evulgare publishes artifacts that are more implementation-shaped:

- Eviulon partnership implementation: https://evulgare.com/docs/implementations/eviulon-defense-partnership
- Eviulon public-source lifecycle: https://evulgare.com/docs/implementations/eviulon-source-lifecycle
- Decision Reconstruction: https://evulgare.com/workbench/decision-reconstruction
- Compartment Continuity: https://evulgare.com/workbench/compartment-continuity
- Sovereign Readiness Evidence Federation: https://evulgare.com/docs/implementations/sovereign-readiness-board
- Confidential Release Review: https://evulgare.com/docs/implementations/confidential-release-review
- Autonomous Defense Authority Kernel: https://evulgare.com/workbench/autonomous-defense-kernel
- Defense Posture: https://evulgare.com/company/defense-posture
- Confidential Defense Assurance: https://evulgare.com/platform/confidential-defense-assurance
- Technical Answerability Federation: https://evulgare.com/workbench/technical-answerability-federation

The public evidence supports several bounded properties: capability is not authority; communications loss does not expand authority; connection restoration does not restore stale authority; reconciliation precedes full control; contradictory evidence remains visible; unattested states are quarantined; readiness gates are conjunctive; and recovery begins from an attested baseline.

The same pages also preserve a critical boundary. Many of these are synthetic, deterministic and non-operational. They use public fixtures or null sinks. They do not demonstrate production control, classified capability, external command execution or independent certification.

## 5. Change detection without invented history

A single capture establishes a baseline. It does not establish a change.

v0.17 therefore labels its external selected-field observations **BASELINE_ESTABLISHED_NO_PRIOR_COMPARABLE_CAPTURE**. Future releases can compare the same selected fields and classify a change. The material dimensions include owner, identifier, authority, jurisdiction, purpose, scope, revision, status, truth boundary, implementation class, operational state, evidence role and restoration/review semantics.

Eviulon's own public drift and correction records can be recorded now as **source-reported events**. Concresca does not convert those first-party reports into an independent diagnosis. The observation registry preserves the origin of the change claim.

## 6. Unknown-preserving results

A field-observation program becomes unreliable when every row must be PASS or FAIL. v0.17 uses a richer result vocabulary:

- supported as first-party doctrine;
- supported in a public implementation record;
- supported in a synthetic demonstration;
- supported in a local design artifact;
- partial;
- not observed;
- contradicted.

The result also states what it does not establish. This is especially important for synthetic artifacts. “Communications loss did not expand authority in the published fixture” is a useful result. “The production system is safe under partition” would be an unsupported upgrade.

## 7. Partition and restoration

Interoperability becomes constitutionally dangerous during loss and recovery. v0.17 formalizes six core rules:

1. Lost connectivity never expands authority.
2. Connection restoration never restores stale authority by itself.
3. Minimum function is bounded and sunsetted.
4. Reconciliation precedes full restoration.
5. Recovery begins from an attested baseline.
6. Restoration remains incomplete until downstream effects are corrected or explicitly listed as residual.

Sixteen analytical state cases test these rules against transport loss, authority expiry, revocation, conflicting histories, attestation mismatch, policy drift, clock divergence, missing restoration acknowledgements, emergency authority, identity-service failure, archive failure, reviewer unavailability and external-source unavailability.

These are test definitions. They report no production recovery time or success rate.

## 8. Human legibility

Constitutional state must be legible to both machines and affected humans. A machine needs stable fields. A person needs a concise answer to: who authorized this, what evidence was used, what is unknown, what changed, what is frozen, what can I do next, and what remains uncorrected?

v0.17 defines fourteen static prototypes: authority receipt, evidence provenance card, uncertainty strip, version/change badge, partition banner, stay/revocation banner, restoration ledger, source-independence label, synthetic/operational badge, reviewer-independence card, reason-code explainer, downstream correction receipt, conjunctive gate panel and unknown-preserving result card.

The prototypes require visible text labels, keyboard-readable structure, plain language, stable identifiers and no color-only meaning. They are not usability-study results or accessibility certification.

## 9. Reconciliation findings

The current public artifacts align on several principles:

- capability does not create authority;
- governance and assurance should remain separate;
- source independence is claim-specific;
- partitions should not expand authority;
- reconnect should not reactivate stale authority;
- synthetic and operational status should remain visible;
- hard gates should not disappear into a composite readiness score.

Other areas remain partial: complete downstream restoration receipts, cross-project appeal/stay propagation, attested reconstitution accepted by all parties, and dual human/machine parity across the full corpus.

Two findings remain deliberately **NOT OBSERVED**: live shared protocol adoption and independent certification.

## 10. What would count as stronger evidence

A future live interoperability observation would need, at minimum:

- identified participating systems and authority owners;
- exact versions and configuration/policy state;
- bounded payload and purpose;
- jurisdiction and authority credentials;
- evidence class and provenance;
- challenge/stay/revocation/restoration semantics;
- actual result and failure state;
- observation period and denominator;
- independent review or reproduction where claimed;
- explicit limitations and residual unknowns.

Without those elements, a public description remains doctrine, implementation evidence or synthetic evidence according to its actual class.

## Final proposition

v0.17 does not claim that Constitutional Interoperability has been adopted. It creates a disciplined way to notice when public artifacts support a property, when they only define it, when they contradict it, and when the answer remains unknown.

The central rule is:

**Observe the artifact. Preserve the source role. Name the property. Record the result class. Publish what is still not established.**