GPT Note — Current Input Truth for Governance Owner Resolution

Date: 2026-06-01

Question

What ensures the system has enough reliable input information to fill governance ownership roles correctly?

Short answer

Currently, the system has many upstream ledgers but not yet enough governed ownership substrate to auto-fill owners safely. It can reliably discover candidates and detect gaps, but it must not auto-assign owner/approval/audit/rollback until SB-1 and SB-2 are built and approved.

Existing sources to reuse

• birth_registry: existence / object birth.
• collection_registry, meta_catalog, table_registry: registry/list/catalog visibility.
• dot_tools, pivot_definitions, event_type_registry, information_unit, iu_relation: domain-specific object sources.
• governance_registry and governance_relations: existing governance agencies and agency-law relations.
• approval_requests, apr_approvals, apr_action_types: approval spine and action vocabulary.
• registry_changelog, event_outbox, system_issues: audit/issue/event substrate.
• label/taxonomy and IU/KG substrates: classification and relation signals.

Current gaps

• No governance_object_ownership table yet.
• No governance_responsibility_scope table yet.
• Missing APR action types for assign_governance_owner, assign_axis_owner, delegate_authority, grant_governance_exception.
• governance_relations only supports law/agency edges, not object/axis ownership.
• No GCOS candidate-state store, source snapshot/ruleset registry, governance worker cursor family, or governance event domain yet.
• IU owner remains unresolved; IU axis envelope still hardcoded at substrate level.

Operational conclusion

Near term: use existing sources to detect candidate objects and gaps. Do not auto-fill ownership as fact. Produce proposals and issues only.

After SB-1/SB-2/GCOS are built: ownership can be filled from governed approval-backed records, not from inference alone.