profiles/corpus-acquisition-policy.v1.yaml

policy_version: corpus-acquisition-policy.v1
status: owner_approved
decision_date: "2026-08-07"
approved_by: owner:chris-page-gov
service_family_denominator: source/service-family-denominator.v1.yaml
reference_inventory: source/exhaustive-reference-inventory.v1.yaml

authority_boundary:
  assertion_status: normalized
  rule: >-
    This policy governs discovery and review. It does not make a service,
    responsibility, provider, geography, legal rule or deadline official.
  prohibited:
    - source_snapshots_without_source_specific_approval
    - source_content_redistribution
    - provider_recommendations_or_rankings
    - inferred_jurisdiction_or_service_equivalence
    - personal_eligibility_legal_or_medical_decisions
    - real_personal_data
    - ci_or_publication_without_explicit_owner_request

local_coverage:
  model: exhaustive_authority_registry_plus_representative_and_exception_leaf_routes
  decision: approved
  authority_layer:
    completeness: every_active_principal_authority_and_functional_body_in_the_approved_official_denominators
    registry: source/authority-registry.v1.yaml
    registry_observed_at: "2026-08-08"
    implementation_status: current_reviewed_denominator_implemented
    sources:
      - govuk-find-local-council
      - mygov-organisations
      - govwales-local-government
      - nidirect-local-councils
      - ons-administrative-names-and-codes
  service_family_layer:
    completeness: every_local_authority_scoped_service_family_mapped_to_responsible_authority_types_in_each_nation
    identity_rule: A responsibility type never proves that a particular authority currently offers the same route.
  leaf_route_layer:
    completeness: one_current_route_per_governed_structural_archetype_plus_every_known_material_exception
    exhaustive_every_authority_required: false
    expansion_rule: A local claim about a named place requires that responsible provider's current leaf source.
  archetypes:
    england:
      - unitary_authority
      - metropolitan_district
      - london_borough
      - county_and_district_two_tier
      - combined_authority_or_mayoral_function
    scotland:
      - council_area
    wales:
      - principal_council
    northern-ireland:
      - district_council
    functional:
      - health_body
      - police_or_fire_area
      - transport_authority
      - court_or_tribunal_region
      - contracted_provider

geography_and_organisation_identifiers:
  decision: approved
  administrative_geography:
    primary_identifier: GSS_nine_character_code
    required_fields:
      - code
      - official_name
      - geography_type
      - vintage_or_effective_date
      - source_url
      - observed_at
    source: ons-administrative-names-and-codes
    history_source: ons-geography-names-codes-lookups
  project_jurisdiction:
    identifiers:
      - uk-wide
      - england
      - scotland
      - wales
      - northern-ireland
      - local
    rule: Project jurisdiction identifiers describe scope and never replace official geography identifiers.
  health_organisations:
    primary_identifier_where_covered: ODS_code
    source: nhs-ods-service
    coverage_source: nhs-ods-data-coverage
    fallback: retain_source_native_identifier_and_identity_evidence
  postcode:
    allowed_use: transient_authority_route_discovery_only
    storage_allowed: false
    identity_rule: A postcode lookup is routing evidence, not a stable service or authority identity.
  lifecycle_rule: Never overwrite predecessor, successor, boundary vintage or effective dates.
  language_variant_identity:
    rule: Treat Welsh and English labels or pages as variants of one identity only when the publisher explicitly pairs them.
    prohibited_evidence:
      - label_similarity
      - translated_slug_similarity
      - parallel_url_pattern_without_publisher_pairing
  local_service_identifiers:
    vocabulary: esd_local_government_services_list
    current_location: https://standards.esd.org.uk/?uri=list%2Fservices
    evidence: https://help.esd.org.uk/standards/faqs/where-can-i-find-latest-local-government-services-list-lgsl
    observed_at: "2026-08-08"
    use: optional_mapping_vocabulary_only
    establishes_current_service_availability: false

health_source_permissions:
  decision: keep_manual_link_and_original_summary_only
  seek_additional_provider_permission_now: false
  reconsider_when: a_specific_approved_acquisition_wave_requires_automated_directory_or_content_extraction
  england:
    source_family: nhs-england-service-directory
    mode: manual_link_and_original_summary_only_despite_ogl
  scotland:
    source_family: nhs-inform-scotland
    mode: manual_link_and_original_summary_only
    reason: personal_noncommercial_terms_and_no_scraping_without_explicit_permission
  wales:
    source_family: nhs-wales-service-directory
    mode: manual_link_and_original_summary_only
    reason: crown_copyright_and_content_authorisation_restrictions
  northern-ireland:
    source_family: hscni-service-directory
    mode: manual_link_and_original_summary_only
    reason: attributed_noncommercial_reproduction_only
  universal_rule: No medical content is transformed into individual advice or eligibility logic.

private_dependencies:
  decision: regulator_first
  provider_selection_rule: >-
    Start with the statutory regulator, mandatory register, official licensing
    authority or statutory redress body. Add a named provider only when the
    journey requires that exact provider and record why it is necessary.
  required_fields:
    - dependency_role
    - sector
    - regulator_or_licensing_authority
    - register_or_directory_url
    - jurisdiction
    - compulsory_or_optional
    - provider_selection_basis
    - observed_at
    - rights_decision
    - redress_route
  prohibited:
    - commercial_rankings
    - inferred_quality_or_endorsement
    - provider_completeness_claim_without_denominator
    - copying_provider_marketing_or_directory_content
  no_single_regulator_rule: Record the absence of a statutory denominator and use a functional role rather than an arbitrary provider.

sector_redress:
  decision: governed_escalation_taxonomy
  registry: source/authority-registry.v1.yaml
  registry_status: ten_sector_regulator_and_redress_map_implemented
  sequence:
    - provider_internal_review_or_complaint
    - statutory_reconsideration_or_review
    - tribunal_or_specialist_appeal
    - regulator_or_ombudsman
    - court_or_judicial_review
  rule: >-
    The sequence is a taxonomy, not a universal procedure. Each service family
    must cite the current route, prerequisites, deadlines and jurisdiction.
  sectors:
    - financial_services_and_insurance
    - housing_and_property
    - energy_and_utilities
    - communications_and_post
    - transport_and_travel
    - health_and_social_care
    - education_and_training
    - legal_and_professional_services
    - employment
    - consumer_goods_and_services

claim_review:
  decision: role_based_nominations
  coordinator:
    reviewer: owner:chris-page-gov
    responsibility: editorial_scope_source_identity_and_release_gate
    status: nominated
  reviewer_roles:
    - id: reviewer:qualified-uk-legal-procedure
      scope: legal_effects_enforcement_appeal_routes_and_statutory_deadlines
      qualification: suitable_current_uk_legal_or_public_policy_expertise
      status: role_nominated_individual_not_appointed
      required_for: production_grade_legal_claim
    - id: reviewer:registered-clinical-safety
      scope: medical_risk_urgent_care_and_health_service_safety_boundaries
      qualification: current_relevant_uk_clinical_registration_or_clinical_safety_expertise
      status: role_nominated_individual_not_appointed
      required_for: production_grade_medical_claim
    - id: reviewer:authoritative-service-policy
      scope: eligibility_evidence_fees_deadlines_and_operational_exceptions
      qualification: current_service_owner_policy_or_equivalent_subject_expertise
      status: assigned_per_service_family
      required_for: high_impact_or_time_critical_claim
  high_impact_rule: >-
    A legal, medical, eligibility, enforcement or deadline claim cannot pass the
    production-grade gate until the applicable role accepts the review and the
    current authoritative source has been re-observed.
  sample_rule: Educational examples may remain clearly labelled normalized or editorial-example with explicit limitations.

explorer_large_corpus:
  target_contract: okf-explorer-large-corpus.v1
  decision: approved_for_local_evaluation
  approval_effective_on: "2026-08-07"
  approval_evidence: evaluation/reviews/integrated-three-slice-2026-08-07.md
  prerequisites_completed:
    - REV-001_closed_relationship_authority_and_provenance
    - REV-002_closed_node_build_provenance
    - REV-003_closed_browser_renderable_source_handoff
    - REV-004_closed_first_class_licence_and_notice_surface
  required_facets:
    - life_course_domain
    - acquisition_wave
    - delivery_scope
    - jurisdiction
    - implementation_status
    - assertion_status
    - rights_state
  publication_allowed: false

link_verification:
  decision: metadata_only_live_audit_separate_from_deterministic_offline_checks
  command: uv run --locked python scripts/audit_source_links.py
  receipt_contract: source-link-receipt.v1
  response_body_retained: false
  offline_build_network_access: false
  primary_failure_effect: blocks_affected_family
  secondary_failure_effect: records_explicit_gap