AZ-052 — Cross-Canon Consistency and Integrity Verification Framework v1
Status
Acest document definește framework-ul formal de verificare a consistenței și integrității între toate canoanele ATLAS ZERO.
După AZ-001 până la AZ-051, există deja:
- specificația protocolului și a subsistemelor;
- canoanele constituționale, parametrice, validatoriale și financiare;
- interfața publică de adevăr al rețelei;
- bundle-ul constituțional machine-readable;
- matricea formală a capabilităților;
- canonul snapshot-urilor istorice publice și regulile lor de verificare în timp.
AZ-052 răspunde la întrebarea: cum verificăm, într-un mod formal, sistematic și machine-readable, că toate canoanele și proiecțiile normative ale protocolului rămân coerente între ele, că nu există drift semantic sau contradicții între identity, parameters, validator set, financial truth, snapshots și exporturile publice și că integritatea sistemului de adevăr al rețelei poate fi validată ca întreg, nu doar pe bucăți izolate?
Scopul documentului este să fixeze:
- modelul verificărilor cross-canon;
- tipurile de consistență și integritate urmărite;
- obiectele de verificare și verdict;
- relația dintre bundle-uri, canoane, snapshot-uri și exporturi;
- regulile de detecție a contradicțiilor, drift-ului și omisiunilor;
- și integrarea acestor verificări în audit, public truth, preservation și governance.
Acest document se bazează pe:
- AZ-002 până la AZ-051, cu accent direct pe AZ-021, AZ-033, AZ-037, AZ-038, AZ-043, AZ-044, AZ-046, AZ-047, AZ-048, AZ-049, AZ-050 și AZ-051.
Termeni:
- MUST = obligatoriu
- MUST NOT = interzis
- SHOULD = recomandat puternic
- MAY = opțional
1. Obiectiv
AZ-052 răspunde la 10 întrebări critice:
- Ce este cross-canon consistency?
- Ce tipuri de contradicții pot exista între canoane?
- Cum verificăm integritatea sistemului de adevăr ca întreg?
- Cum legăm boundary-ul, branch-ul și identity-ul între canoane?
- Cum detectăm drift-ul dintre canoane și proiecții publice?
- Cum distingem inconsistență reală de diferență legitimă de scope?
- Cum modelăm verdictul de consistență și severitatea lui?
- Cum integrăm aceste verificări în audit, public truth și preservation?
- Cum tratăm istoricul inconsistențelor, remedierilor și supersession-ului?
- Cum evităm ca fiecare canon să fie corect local, dar sistemul global să fie incoerent?
2. Principii
2.1 Local validity is not enough
Faptul că fiecare canon este local bine format nu este suficient. Sistemul MUST verifica și coerența dintre ele.
2.2 Scope mismatch must be distinguished from contradiction
Două obiecte pot fi diferite legitim dacă au scope sau boundary diferit. Framework-ul MUST separa diferența legitimă de contradicția nelegitimă.
2.3 Consistency checks must be explicit and typed
Verificările cross-canon SHOULD fi tipizate, reproducibile și machine-readable, nu inferate vag din documentație.
2.4 Integrity means both completeness and coherence
Integritatea globală înseamnă:
- lipsa coruperii;
- lipsa omisiunilor materiale;
- și lipsa contradicțiilor semantice între canoane.
2.5 Public projections must remain subordinate to source canons
Exporturile și proiecțiile publice MUST rămână consistente cu canoanele sursă și MUST NOT deveni surse paralele de adevăr.
2.6 Historical and branch-aware verification is mandatory
Verificările de integritate SHOULD funcționa nu doar pentru starea curentă, ci și pentru snapshot-uri istorice și ramuri constituționale distincte.
3. Framework purpose
3.1 Framework-ul SHOULD răspunde:
- aceste canoane se referă la aceeași rețea și același branch?
- același boundary este reprezentat coerent în toate bundle-urile?
- validator set-ul este compatibil cu parameter set-ul și cu constitutional scope?
- financial truth este coerent cu validator truth și parameter truth?
- public truth projection corespunde surselor canonice?
- snapshot-ul istoric folosește schema și semantica potrivite?
- există drift, contradicții sau omisiuni materiale?
3.2 Rule
Dacă nu poate răspunde clar la aceste întrebări, framework-ul este insuficient.
4. Canon classes in scope
4.1 Framework-ul SHOULD acoperi cel puțin:
- constitutional record canon
- protocol parameter registry canon
- validator membership canon
- public financial truth canon
- public network truth interface
- machine-readable constitution bundle
- public historical snapshot canon
- capability matrix where relevant
- audit export projections where relevant
- preservation records where relevant
4.2 Rule
No canon with normative truth significance SHOULD remain outside consistency framework dacă influențează restul sistemului.
5. Consistency dimensions
5.1 Recommended dimensions
ATLAS ZERO SHOULD verifica cel puțin:
- identity consistency
- branch consistency
- boundary consistency
- parameter-validator consistency
- parameter-financial consistency
- validator-financial consistency
- projection-source consistency
- historical consistency
- supersession consistency
- schema and verification-profile consistency
- archive/preservation consistency
- capability-authority consistency
5.2 Rule
Every major cross-canon relationship SHOULD map to at least one explicit consistency dimension.
6. Cross-canon verification scope model
6.1 Canonical structure
CrossCanonVerificationScope {
verification_scope_id
target_network_class
target_chain_id
target_genesis_hash
constitutional_branch_ref?
target_boundary_hash
included_canon_root
included_projection_root?
}
6.2 Rule
Cross-canon verification MUST be scope-bound and MUST state which canons/projections are being compared.
7. Verification relation model
7.1 Need
Each consistency check compares one or more sources under a specific relation.
7.2 Canonical structure
CrossCanonRelation {
relation_id
relation_class
left_ref_root
right_ref_root
relation_scope_hash
}
7.3 relation_class examples
- same_identity_expected
- same_boundary_expected
- same_branch_expected
- derived_projection_expected
- subset_consistency_expected
- lineage_continuity_expected
- authority_scope_expected
7.4 Rule
Relations MUST be explicit so that a verifier knows what is supposed to match.
8. Consistency check classes
8.1 Recommended check classes
ATLAS ZERO SHOULD include cel puțin:
CCC_IDENTITY_MATCHCCC_BRANCH_MATCHCCC_BOUNDARY_MATCHCCC_CANON_ROOT_ALIGNMENTCCC_PARAMETER_SET_ALIGNMENTCCC_VALIDATOR_SET_ALIGNMENTCCC_FINANCIAL_STATE_ALIGNMENTCCC_PROJECTION_ALIGNMENTCCC_HISTORICAL_SNAPSHOT_ALIGNMENTCCC_SUPERSESSION_LINEAGE_ALIGNMENTCCC_SCHEMA_PROFILE_ALIGNMENTCCC_PRESERVATION_ALIGNMENTCCC_CAPABILITY_SCOPE_ALIGNMENT
8.2 Rule
Check classes MUST be stable enough for historical trend analysis and audit.
9. Consistency check object
9.1 Canonical structure
CrossCanonConsistencyCheck {
version_major
version_minor
check_id
check_class
verification_scope_ref
relation_ref
expected_invariant_hash
evidence_root?
severity_class
}
9.2 severity_class
- informational
- caution
- material
- critical
- constitutional
9.3 Rule
Every check MUST define the invariant it expects, not merely the data it inspects.
10. Integrity check classes
10.1 Integrity extends beyond equality and includes:
- required canon present
- required bundle present
- required historical lineage preserved
- required supersession records present
- required schema metadata present
- required verification profile present
- required archive linkage present
10.2 Recommended integrity classes
CCI_REQUIRED_COMPONENT_PRESENTCCI_REQUIRED_LINEAGE_PRESENTCCI_REQUIRED_SCHEMA_PRESENTCCI_REQUIRED_PROFILE_PRESENTCCI_REQUIRED_ARCHIVE_LINK_PRESENTCCI_REQUIRED_PROJECTION_LINK_PRESENT
10.3 Rule
Integrity checks MUST catch omission, not only contradiction.
11. Invariant model
11.1 An invariant is the precise condition that must hold across multiple canons or bundles.
11.2 Canonical structure
CrossCanonInvariant {
invariant_id
invariant_class
statement_hash
scope_hash
allowed_variation_root?
}
11.3 invariant_class examples
- exact_root_match
- exact_scope_match
- branch_compatible_match
- historical_boundary_equivalence
- projection_is_derived_subset
- supersession_chain_contiguous
- parameter_truth_matches_financial_rules
11.4 Rule
Allowed variation MUST be explicit when exact equality is not required.
12. Allowed variation model
12.1 Some differences are legitimate, for example:
- public projection includes subset only
- historical snapshot uses branch-specific completeness
- audit export includes extra evidence
- verifier profile is adapted but semantically equivalent
12.2 Canonical structure
AllowedVariationRule {
variation_rule_id
invariant_ref
variation_class
condition_hash
}
12.3 variation_class examples
- subset_projection_allowed
- branch_specific_divergence_allowed
- historical_profile_adapter_allowed
- archival_fallback_allowed
12.4 Rule
No legitimate variation SHOULD remain implicit.
13. Identity consistency
13.1 Identity consistency SHOULD verify that:
- constitutional canon
- parameter canon
- validator canon
- financial canon
- public truth projection
- constitution bundle all bind the same:
- target_network_class
- chain_id
- genesis_hash
- branch reference for the declared scope.
13.2 Rule
Identity mismatch is one of the strongest consistency failures.
14. Boundary consistency
14.1 Boundary consistency SHOULD verify that:
- active/historical boundary semantics match across canons;
- parameter set boundary aligns with validator set boundary;
- financial state boundary aligns with constitutional/public snapshot boundary;
- exported summary does not silently shift boundary.
14.2 Rule
Boundary ambiguity SHOULD be treated as material or critical depending on scope.
15. Branch consistency
15.1 Branch consistency SHOULD verify that:
- branch refs are present where required;
- branch-specific canons do not claim same-chain equivalence improperly;
- snapshots and public exports identify the correct branch;
- validator and financial truth are branch-local where necessary.
15.2 Rule
Flattening branch-specific truth into one canonical view after split is invalid.
16. Parameter-validator consistency
16.1 This dimension SHOULD verify:
- validator set size/shape is valid under active parameter rules;
- role assignment assumptions match parameterized thresholds;
- upgrade or recovery constraints on validators align with active parameter scope;
- public projections do not expose stale validator truth under new parameters.
16.2 Rule
A validator set inconsistent with active parameter truth is a serious integrity failure.
17. Parameter-financial consistency
17.1 This dimension SHOULD verify:
- fees, rewards, slashing and reserve semantics implied by financial state match active parameters;
- public obligations are compatible with active rules;
- branch/recovery-specific financial flows are compatible with active parameter boundary.
17.2 Rule
Financial truth without parameter compatibility is semantically unstable.
18. Validator-financial consistency
18.1 This dimension SHOULD verify:
- reward eligibility matches legitimate validator membership;
- slashing-related financial flows map to validator legitimacy changes where policy requires;
- revoked or branch-excluded validators do not appear in incompatible public financial obligations.
18.2 Rule
Validator and financial canons MUST be joined coherently whenever rewards or penalties are in scope.
19. Projection-source consistency
19.1 This dimension SHOULD verify:
- public truth exports match source canons for overlapping claims;
- audit exports do not contradict public exports for same scope;
- public historical snapshots remain traceable to boundary-specific constitution bundles;
- minimal projections do not overclaim completeness.
19.2 Rule
Projection-source drift SHOULD be treated as material or critical depending on claim class.
20. Historical consistency
20.1 Historical consistency SHOULD verify:
- snapshot boundary is correctly pinned;
- historical branch classification is stable;
- verification profile matches snapshot schema;
- source constitution bundle or source roots match the claimed boundary;
- superseded historical snapshots remain linked and not silently replaced.
20.2 Rule
Historical truth MUST be protected from retroactive reinterpretation.
21. Supersession consistency
21.1 This dimension SHOULD verify:
- supersession chains are contiguous;
- superseding bundles or snapshots do not erase prior discoverability;
- branch-specific supersession is not mistaken for global supersession;
- revocation and supersession semantics are not mixed improperly.
21.2 Rule
Broken supersession lineage is an integrity defect.
22. Schema and verification-profile consistency
22.1 This dimension SHOULD verify:
- schema versions match included records;
- verifier metadata and profiles correspond to the snapshot or export they govern;
- archival fallback semantics are declared when direct verification is no longer native;
- schema upgrades do not silently change historical interpretation.
22.2 Rule
Schema/profile mismatches can invalidate otherwise correct data.
23. Preservation consistency
23.1 This dimension SHOULD verify:
- preserved archives correspond to claimed canon roots;
- historical snapshots preserved are the same ones referenced publicly;
- constitution bundles, public projections and archive manifests remain linked;
- verification runs correspond to actual preserved artifacts.
23.2 Rule
Preservation drift SHOULD be visible in consistency framework, not only in storage tooling.
24. Capability-authority consistency
24.1 This dimension SHOULD verify:
- actor classes authorized to attest or mutate truth match the canons they affect;
- public/verifier actors do not appear as normative authorities unless explicitly granted;
- bundle projections and exports are emitted by actors with correct capability grants.
24.2 Rule
Authority mismatch is both governance and integrity failure.
25. Cross-canon verification run object
25.1 Canonical structure
CrossCanonVerificationRun {
version_major
version_minor
run_id
verification_scope_ref
check_root
integrity_check_root?
started_at_unix_ms
completed_at_unix_ms?
result_root
result_class
notes_hash?
}
25.2 result_class
- pass
- pass_with_notes
- partial
- fail_material
- fail_critical
- fail_constitutional
25.3 Rule
A serious system state SHOULD periodically or event-driven produce verification runs.
26. Check result model
26.1 Canonical structure
CrossCanonCheckResult {
result_id
check_ref
verdict_class
observed_root?
contradiction_root?
notes_hash?
}
26.2 verdict_class
- satisfied
- satisfied_with_allowed_variation
- inconclusive
- violated_material
- violated_critical
- violated_constitutional
26.3 Rule
Results MUST distinguish allowed variation from true contradiction.
27. Contradiction model
27.1 Need
A contradiction is a material failure of an invariant, not a mere difference.
27.2 Canonical structure
CrossCanonContradictionRecord {
contradiction_id
contradiction_class
left_ref_root
right_ref_root
violated_invariant_ref
severity_class
evidence_root?
timestamp_unix_ms
}
27.3 contradiction_class examples
- identity_conflict
- boundary_conflict
- branch_conflict
- parameter_financial_conflict
- validator_membership_conflict
- projection_source_conflict
- schema_interpretation_conflict
- supersession_lineage_break
27.4 Rule
Contradictions SHOULD be preserved and queryable historically.
28. Omission model
28.1 Need
A canon ecosystem can fail by missing required pieces, not only by contradiction.
28.2 Canonical structure
CrossCanonOmissionRecord {
omission_id
omission_class
missing_component_class
affected_scope_hash
severity_class
timestamp_unix_ms
}
28.3 omission_class examples
- missing_required_canon
- missing_required_projection_link
- missing_required_historical_snapshot
- missing_schema_metadata
- missing_preservation_link
- missing_supersession_record
28.4 Rule
Omissions SHOULD feed integrity results and remediation.
29. Drift model
29.1 Drift means semantically relevant divergence between intended aligned sources over time.
29.2 Drift MAY occur through:
- stale public projections
- stale verifier metadata
- outdated snapshot indexing
- public truth export not refreshed after branch or upgrade
- archive references lagging behind active canon
29.3 Canonical structure
CrossCanonDriftRecord {
drift_id
drift_class
source_ref
derived_ref
expected_alignment_ref
severity_class
timestamp_unix_ms
}
29.4 Rule
Drift SHOULD be explicitly detected, not only guessed from inconsistent UX.
30. Remediation model
30.1 Integrity issues SHOULD lead to remediations such as:
- refresh public projection
- reissue historical snapshot with supersession
- repair bundle linkage
- add missing schema/profile metadata
- correct branch classification
- re-run preservation verification
- open governance review for constitutional conflict
30.2 Canonical structure
CrossCanonRemediationRecord {
remediation_id
source_issue_ref
remediation_class
owner_role_class
due_boundary_hash?
status
}
30.3 status
- proposed
- approved
- in_progress
- implemented
- verified
- deferred
30.4 Rule
Critical contradictions SHOULD have tracked remediation, not only logging.
31. Trigger model
31.1 Cross-canon verification SHOULD run:
- on major launch boundary
- on upgrade activation
- on hard fork or branch split
- on recovery ratification
- on public truth export issuance
- on historical snapshot publication
- on preservation verification change
- on governance acts affecting constitutional truth
31.2 Rule
Periodic cadence MAY also exist, but boundary/event triggers SHOULD be primary for correctness-sensitive scopes.
32. Trigger record
32.1 Canonical structure
CrossCanonVerificationTrigger {
trigger_id
trigger_class
related_ref
target_scope_hash
timestamp_unix_ms
}
32.2 trigger_class examples
- launch_boundary
- upgrade_boundary
- fork_boundary
- recovery_boundary
- snapshot_publish
- public_projection_publish
- archive_verification_change
- governance_constitutional_act
32.3 Rule
Triggers SHOULD be archived for auditability of why a run happened.
33. Relationship to public truth interface
33.1 PNTI SHOULD expose enough metadata to support:
- projection-source consistency checks
- current/historical branch alignment checks
- verification-profile alignment for historical snapshots
- supersession lineage checks
33.2 Rule
Public truth integrity SHOULD be subset-compatible with full cross-canon framework.
34. Relationship to constitution bundle
34.1 The machine-readable constitution bundle SHOULD either:
- embed selected consistency assertions;
- or point to current cross-canon verification results for its scope.
34.2 Rule
A current-active constitution bundle SHOULD not exist in isolation from integrity verification.
35. Relationship to audit exports
35.1 Audit exports MAY include:
- relevant verification runs
- contradiction records
- omission records
- drift records
- remediation status
- cross-canon assertion subsets
35.2 Rule
Where audit claims depend on whole-system integrity, cross-canon evidence SHOULD be exportable.
36. Relationship to preservation and historical verification
36.1 Historical bundles and snapshots SHOULD be rechecked against:
- preserved schema metadata
- preserved source canon roots
- preserved supersession lineage
- preserved branch identity
36.2 Rule
Cross-canon consistency MUST survive time, not only current runtime.
37. Relationship to capability matrix
37.1 Capability matrix SHOULD constrain:
- who can issue projections;
- who can attest consistency runs;
- who can remediate contradictions;
- who can approve constitutional corrections.
37.2 Rule
Consistency framework MUST not imply authority that the capability matrix does not grant.
38. Query model
38.1 Useful queries
- is current public truth projection aligned with active constitution bundle?
- did upgrade boundary X introduce any cross-canon contradictions?
- which historical snapshots are currently archival-only verifiable?
- are validator and financial canons aligned for branch B at boundary Y?
- what contradictions remain unresolved?
- what bundle supersession chains are broken or incomplete?
38.2 Rule
Framework SHOULD support issue-oriented and scope-oriented queries.
39. Anti-patterns
39.1 Systems SHOULD avoid:
- validating each canon separately and assuming system integrity
- no distinction between contradiction and allowed variation
- branch-specific mismatch hidden as stale export
- public projections refreshed without checking source roots
- historical snapshot verified under current rules only
- missing bundle lineage treated as harmless
- integrity checks defined only in prose
- no omission detection
- no remediation tracking for critical contradictions
- no preserved history of resolved contradictions and drift
40. Formal goals
AZ-052 urmărește aceste obiective:
40.1 Whole-system integrity
The protocol can verify the coherence of its truth system as a whole.
40.2 Typed contradiction detection
Contradictions, omissions and drift become machine-detectable and classifiable.
40.3 Historical and branch-aware integrity
Integrity verification remains meaningful across time, upgrades, recovery and branch splits.
40.4 Audit and governance usability
Cross-canon verification results can support audits, public exports, remediation and constitutional decisions.
41. Formula documentului
Cross-Canon Integrity = scope-bound verification runs + typed invariants + consistency and integrity checks + contradiction/omission/drift records + remediation lineage
42. Relația cu restul suitei
- AZ-048 definește proiecția publică a adevărului.
- AZ-049 definește bundle-ul constituțional machine-readable.
- AZ-051 definește snapshot-urile istorice publice.
- AZ-052 definește verificarea transversală care le ține coerente împreună.
Pe scurt: AZ-052 este stratul de integritate globală al întregului sistem de canoane și proiecții normative.
43. Ce urmează
După AZ-052, documentul corect este:
AZ-053 — Protocol Documentation Closure and Canon Index
Acolo trebuie fixate:
- indexul final al întregii suite;
- relațiile dintre toate documentele și canoanele;
- ce este normativ, operațional, auditabil sau public;
- și închiderea formală a corpusului într-o structură navigabilă și verificabilă ca ansamblu.
Închidere
Un protocol matur nu este doar un set de canoane bine scrise. Este un sistem care poate demonstra că aceste canoane nu se contrazic, că proiecțiile lor rămân fidele surselor, că istoria lor rămâne coerentă în timp și că orice drift sau ruptură devine vizibilă înainte să devină adevăr greșit.
Acolo începe verificarea reală a integrității globale.