# Vercy AI instruction - YAML 1.2 (JSON-compatible) { "vercy": "1.0-draft", "publication": { "status": "published", "adjudicationStatus": "reviewable-draft", "publishableCanonical": false, "generatedAt": "2026-09-02T19:05:14Z", "synthesisSha256": "881bf3f867b0c3c6ee5ba7630e51b942d2a16e135600f67ffddb3094c82ba374", "providerMode": "single-provider-waiver", "providers": [ "Claude" ], "waivedProviders": [ "Grok" ] }, "metaModel": { "id": "WM-KNW-010", "registryId": "vr.wm-knw-010", "name": "Decision / Rationale", "version": "0.3.0-research.1", "previousVersions": [], "entryKind": "aggregate", "family": "World Models", "category": "Information and virtual systems", "industry": [ "Cross-industry" ], "domain": [ "INF.KNW.DEC" ], "tags": [ "decision", "rationale", "inf.knw.dec" ], "status": "published" }, "canonicalUrl": "https://ver.cy/models/wm-knw-010-decision-rationale/", "sourceUrl": "https://github.com/ver-cy/world-models/tree/feat/mega-model-registry/research/runs/wm-knw-010", "model": { "registry_id": "vr.wm-knw-010", "model_id": "WM-KNW-010", "name": "Decision / Rationale", "entry_kind": "aggregate", "purpose": "Govern the decision as an inspectable knowledge object: the issue to be resolved, the objectives and constraints that frame it, the alternatives considered, the criteria and evaluation applied, the selected outcome and its disposition, together with the rationale, evidence citations, authority, lifecycle, provenance and access governance that let a machine agent understand, reconstruct and reuse the choice.", "scope_statement": "WM-KNW-010 governs the content of a decision taken on a specific occasion and the reasoning offered for it, independently of storage format and access interface. It covers stable decision identity and version discrimination; decision type and jurisdictional profile; the decision question, objectives, context and affected subjects; constraints and the decision rules relied upon; the candidate alternative set including any required no-action baseline; eligibility screening and recorded exclusion; criteria, measures, scales and weights; evaluation results, comparison and sensitivity; the selected alternative, disposition, explicit outcome statement, attached conditions and stated consequences; the rationale and argument structure that justify the outcome together with citations to evidence; the participants, roles and authority basis under which the decision was taken; and the lifecycle, provenance, access, retention and quality governance of decision content. It does not govern the activity that produced the decision, the document that carries it, the legal instrument that may enact it, the evidence items it cites, any executable rule or decision service, or downstream implementation.", "in_scope": [ "Stable decision identity, version discriminator and supersession links between decisions", "Decision type, jurisdictional profile binding and profile-required content elements", "The decision question, its admissible answer space and the problem or issue that triggered it", "Objectives, critical success factors and expected effects on declared indicators", "Decision context, dependencies on input data and upstream decisions, and affected subjects", "Binding constraints, admissibility conditions and the decision rules relied upon, stated declaratively", "The registered candidate alternative set, including a no-action or business-as-usual baseline where a profile requires one", "Eligibility screening and recorded reasons for excluding alternatives from detailed evaluation", "Evaluation criteria, measures, scales, preference directions and weighting schemes", "Recorded evaluation results, aggregation, ranking and sensitivity findings", "The selected alternative, disposition, explicit outcome statement, effect onset and applicability", "Conditions, reservations, accepted risks, stated consequences and follow-on decisions", "Rationale, argument structure and explanation linking the outcome to objectives, criteria and cited evidence", "Participants, roles, delegation and the authority basis on which the decision was taken", "Decision-content lifecycle states, revision history, provenance, access classification, retention and quality governance" ], "out_of_scope": [ "The decision-making or approval activity itself, including convening, deliberation, voting and approval execution (WM-ACT-024)", "The document or record container that renders, files, publishes or preserves the decision (WM-REC-010)", "The legal judgment as a binding legal instrument, its enforceability and its citation regime (WM-POL-021)", "Generic claim assertion, truth-value and claim-conflict machinery (WM-KNW-007)", "Evidence items themselves, their appraisal, integrity and chain of custody (WM-KNW-008)", "Executable rules, policies, decision services and rule engines, and their runtime evaluation or enforcement", "Audit-trail records produced by an external audit or logging service", "Implementation, delivery and change execution that follows a decision (WM-ACT-032)", "Domain-specific decision rule content such as clinical protocols, credit scorecards, procurement award formulas or sentencing guidelines", "Mathematical decision-analysis method definitions such as utility theory, AHP, ELECTRE or PROMETHEE, which bind only as optional external methods" ], "boundary_notes": [ { "neighbor": "WM-ACT-024 Decision-making / approval activity", "distinction": "This model governs decision content. DMN's Decision element only points at decisionMaker, decisionOwner, usingProcess and usingTask by reference, and ISO/IEC/IEEE 15288 places decision management as a process. Convening, deliberation, voting and approval execution therefore belong to the activity model and are referenced, never reproduced here.", "source_refs": [ "SRC-001", "SRC-004" ] }, { "neighbor": "WM-REC-010 Document / record container", "distinction": "Akoma Ntoso separates the abstract Work and its Expression from the Manifestation and Item that carry it. Decision content is the governed subject here; rendering, filing, formatting, publication and preservation of the carrying record belong to the record model.", "source_refs": [ "SRC-012" ] }, { "neighbor": "WM-POL-021 Legal judgment", "distinction": "A judgment is a legal instrument with binding force, its own citation regime and a judgmentBody whose introduction, background, motivation and decision sections are editorially structured. This model captures the decision content a judgment may express, not the instrument's legal effect.", "source_refs": [ "SRC-012", "SRC-005" ] }, { "neighbor": "WM-KNW-007 Claim (registry parent)", "distinction": "A claim asserts that something is the case. A decision selects among registered alternatives and produces a disposition. Generic assertion identity, truth-value and claim-conflict machinery stay with the parent model; only decision-specific specialisation is modelled here.", "source_refs": [ "SRC-003", "SRC-012" ] }, { "neighbor": "WM-KNW-008 Evidence", "distinction": "Cited evidence is bound by identifier with a stated role in an evaluation result. Evidence appraisal, integrity, sufficiency and chain of custody remain with the evidence model; NASA's guidance on assumptions and their supporting evidence is used only to justify the citation binding.", "source_refs": [ "SRC-008", "SRC-005" ] }, { "neighbor": "Executable rule, policy or decision service", "distinction": "DMN decision logic, business knowledge models and decision tables are executable artefacts deployed and run by a decision service. This model records the rules relied upon, their declared hit policy and a versioned reference to the source rule, and never owns runtime evaluation, enforcement or the engine's execution trace.", "source_refs": [ "SRC-001", "SRC-002" ] }, { "neighbor": "WM-ACT-032 Change request", "distinction": "A change request may be resolved by a decision; the decision carries a back-reference to the resolved issue. The request's own lifecycle, prioritisation and the implementation of the change belong to that model, as does the monitoring and compliance work that a record of decision may direct.", "source_refs": [ "SRC-010", "SRC-004" ] }, { "neighbor": "WM-KNW-010 rationale, participation and governance areas (same model)", "distinction": "Argument structure, explanation, participants and authority basis, and the lifecycle, provenance, access, retention and audit governance of decision content are governed by the other areas of this same model. The decision-core structures supply their subject matter by reference and do not restate them.", "source_refs": [ "SRC-003", "SRC-004" ] } ] }, "sources": [ { "id": "SRC-001", "title": "DMN 1.5 XML Schema (DMN15.xsd) — normative machine-readable file", "organization": "Object Management Group (OMG)", "url": "https://www.omg.org/spec/DMN/20230324/DMN15.xsd", "version_or_date": "DMN 1.5 machine-readable file, namespace dated 2023-03-24, issued with formal/24-01-01", "source_type": "schema", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T09:12:00Z", "relevance": "Normative element definitions: tDecision carries question, allowedAnswers, variable, informationRequirement, knowledgeRequirement, authorityRequirement, supportedObjective, impactedPerformanceIndicator, decisionMaker, decisionOwner, usingProcess, usingTask and an expression; tDecisionTable carries hitPolicy, aggregation, input, output and rule; tDMNElement carries id, label and description." }, { "id": "SRC-002", "title": "About the Decision Model and Notation Specification Version 1.5", "organization": "Object Management Group (OMG)", "url": "https://www.omg.org/spec/DMN/1.5/About-DMN", "version_or_date": "Version 1.5, August 2024, document number formal/24-01-01", "source_type": "standard", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T09:10:00Z", "relevance": "Establishes the formal identity, version and date of the DMN specification used as the alignment target, and lists its normative machine-readable files." }, { "id": "SRC-003", "title": "IEEE/ISO/IEC 42010-2022 — International Standard for Software, systems and enterprise: Architecture description", "organization": "IEEE Standards Association with ISO and IEC", "url": "https://standards.ieee.org/ieee/42010/6846/", "version_or_date": "IEEE/ISO/IEC 42010-2022, published 2022-11-07, board approval 2022-09-21", "source_type": "standard", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T09:16:00Z", "relevance": "Treats architecture decision and architecture rationale as distinct first-class constructs, relates decisions to concerns and to the elements they affect, and notes that making a decision can raise new concerns; supports separating decision content from rationale and grounds the consequences finding." }, { "id": "SRC-004", "title": "IEEE/ISO/IEC 15288-2023 — Systems and software engineering: System life cycle processes", "organization": "IEEE Standards Association with ISO and IEC", "url": "https://standards.ieee.org/ieee/15288/10424/", "version_or_date": "IEEE/ISO/IEC 15288-2023, published 2023-05-16, board approval 2023-03-30", "source_type": "standard", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T09:18:00Z", "relevance": "Establishes decision management as a technical management process within the system life cycle, which is the basis for placing the decision-making activity outside this content model while keeping the decision's own content in scope." }, { "id": "SRC-005", "title": "Regulation (EU) 2016/679 (General Data Protection Regulation)", "organization": "European Union (published via EUR-Lex)", "url": "https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:32016R0679", "version_or_date": "Adopted 27 April 2016, CELEX 32016R0679", "source_type": "legislation", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T09:06:00Z", "relevance": "Article 22 defines decisions based solely on automated processing that produce legal or similarly significant effects for a data subject; Article 15(1)(h) requires meaningful information about the logic involved. Grounds the decision-type and jurisdictional-profile finding, the affected-subjects element and the explainability requirement on evaluation content." }, { "id": "SRC-006", "title": "The Green Book (2026): appraisal and evaluation in central government", "organization": "HM Treasury (United Kingdom)", "url": "https://www.gov.uk/government/publications/the-green-book-appraisal-and-evaluation-in-central-government/the-green-book-2026", "version_or_date": "2026 edition, updated 5 February 2026", "source_type": "public-authority", "primary_source": true, "authority_tier": 2, "accessed_at": "2026-09-02T09:24:00Z", "relevance": "Requires a case for change, SMART objectives, critical success factors, a broad longlist narrowed to a shortlist, and a mandatory business-as-usual benchmark option alongside a do-minimum option; sets out appraisal criteria, monetisable and non-monetisable impacts, optimism-bias adjustment, and an appraisal summary table recording each option's results and the preferred option." }, { "id": "SRC-007", "title": "Use of Multi-Criteria Decision Analysis in options appraisal of economic cases (Green Book supplementary guidance)", "organization": "HM Treasury (United Kingdom)", "url": "https://www.gov.uk/government/publications/green-book-supplementary-guidance-multi-criteria-decision-analysis/use-of-multi-criteria-decision-analysis-in-options-appraisal-of-economic-cases", "version_or_date": "Published 16 May 2024", "source_type": "public-authority", "primary_source": true, "authority_tier": 2, "accessed_at": "2026-09-02T09:26:00Z", "relevance": "Defines criteria derivation from critical success factors, measurability and mutual independence of criteria, reference-scale scoring, swing weighting, sensitivity analysis on scores and weights, and the requirement to document results transparently; also cautions that MCDA suits longlist appraisal only." }, { "id": "SRC-008", "title": "NASA Systems Engineering Handbook, Section 6.8 Decision Analysis", "organization": "National Aeronautics and Space Administration (NASA)", "url": "https://www.nasa.gov/reference/6-8-decision-analysis/", "version_or_date": "NASA Systems Engineering Handbook, section last updated 26 July 2023", "source_type": "public-authority", "primary_source": true, "authority_tier": 2, "accessed_at": "2026-09-02T09:34:00Z", "relevance": "Sets out decision framing, objective and measurable criteria that differentiate alternatives, mandatory versus enhancing criteria, alternative generation and trade-tree pruning, early screening of options failing mandatory criteria, normalisation for comparison, sensitivity and uncertainty analysis, and decision report contents including criteria, alternatives, assumptions, results, recommendation and the final decision with rationale." }, { "id": "SRC-009", "title": "40 CFR 1502.14 — Alternatives including the proposed action (2024 CFR annual edition)", "organization": "U.S. Government Publishing Office, for the Council on Environmental Quality", "url": "https://www.govinfo.gov/content/pkg/CFR-2024-title40-vol37/xml/CFR-2024-title40-vol37-sec1502-14.xml", "version_or_date": "CFR 2024 annual edition, title 40, section 1502.14 (removed effective 11 April 2025)", "source_type": "legislation", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T09:20:00Z", "relevance": "Requires rigorous exploration and objective evaluation of reasonable alternatives, inclusion of the no action alternative, identification of the preferred alternative in the draft statement and identification of the environmentally preferable alternative. Cited as a historical normative content pattern for the mandatory baseline and for recorded exclusion of alternatives; see SRC-011 for its removal." }, { "id": "SRC-010", "title": "40 CFR 1505.2 — Record of decision in cases requiring environmental impact statements (2024 CFR annual edition)", "organization": "U.S. Government Publishing Office, for the Council on Environmental Quality", "url": "https://www.govinfo.gov/content/pkg/CFR-2024-title40-vol37/xml/CFR-2024-title40-vol37-sec1505-2.xml", "version_or_date": "CFR 2024 annual edition, title 40, section 1505.2 (removed effective 11 April 2025)", "source_type": "legislation", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T09:22:00Z", "relevance": "Requires a decision record to state the decision, identify the alternatives considered and which were environmentally preferable, identify and discuss all factors including essential national policy considerations that were balanced, and state whether all practicable mitigation was adopted and if not why, with a monitoring and compliance plan where mitigation is adopted. Grounds the outcome, balanced-factors and attached-conditions findings; see SRC-011 for its removal." }, { "id": "SRC-011", "title": "Removal of National Environmental Policy Act Implementing Regulations (interim final rule)", "organization": "Council on Environmental Quality (United States)", "url": "https://www.govinfo.gov/content/pkg/FR-2025-02-25/html/2025-03014.htm", "version_or_date": "90 FR 10610, published 25 February 2025, effective 11 April 2025", "source_type": "public-authority", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T09:40:00Z", "relevance": "Removes 40 CFR parts 1500 to 1508 and directs agencies to rely on their own NEPA implementing procedures. Recorded as a conflict so that the alternatives and record-of-decision content patterns are cited as historical and jurisdiction-specific rather than as currently binding government-wide requirements." }, { "id": "SRC-012", "title": "Akoma Ntoso Version 1.0 Part 1: XML Vocabulary", "organization": "OASIS", "url": "https://docs.oasis-open.org/legaldocml/akn-core/v1.0/os/part1-vocabulary/akn-core-v1.0-os-part1-vocabulary.html", "version_or_date": "OASIS Standard, 29 August 2018", "source_type": "standard", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T09:30:00Z", "relevance": "Applies the FRBR Work, Expression, Manifestation and Item levels to persistent identification of legal content, and structures a judgment body into introduction, background, motivation and decision. Grounds the separation of decision content from its record container, the content-versus-container identity distinction, and the boundary with a legal judgment." }, { "id": "SRC-013", "title": "The Argument Interchange Format (AIF) Specification", "organization": "ARG-tech / AIF Technical Committee", "url": "https://www.arg-tech.org/wp-content/uploads/2011/09/aif-spec.pdf", "version_or_date": "Draft specification, document dated 2011-08-29", "source_type": "ontology", "primary_source": true, "authority_tier": 3, "accessed_at": "2026-09-02T09:06:00Z", "relevance": "Defines the core argumentation ontology: information nodes, rule-application (inference), conflict-application and preference-application nodes, and the Forms Ontology of schemes with premises, conclusions, exceptions and presumptions. Grounds the argument-structure and inference-scheme findings." }, { "id": "SRC-014", "title": "PROV-O: The PROV Ontology", "organization": "World Wide Web Consortium (W3C)", "url": "https://www.w3.org/TR/prov-o/", "version_or_date": "W3C Recommendation, 30 April 2013 (REC-prov-o-20130430)", "source_type": "standard", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T09:06:00Z", "relevance": "Supplies Entity, Activity, Agent and SoftwareAgent with wasAttributedTo, wasGeneratedBy, wasDerivedFrom, wasAssociatedWith, qualifiedAttribution and generatedAtTime. Grounds reasoning provenance and the authored-versus-generated distinction." }, { "id": "SRC-015", "title": "Structured Assurance Case Metamodel (SACM), Version 2.3", "organization": "Object Management Group (OMG)", "url": "https://www.omg.org/spec/SACM/2.3/About-SACM", "version_or_date": "Version 2.3, adopted October 2023", "source_type": "standard", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T09:09:00Z", "relevance": "Defines an assurance case as auditable claims, arguments and evidence, with Claim, ArgumentReasoning, AssertedInference and AssertedEvidence links plus explicit counter-argument and counter-evidence. Grounds argument structure, claim and evidence bindings and contestation." }, { "id": "SRC-016", "title": "NISTIR 8312: Four Principles of Explainable Artificial Intelligence", "organization": "National Institute of Standards and Technology (NIST), U.S. Department of Commerce", "url": "https://nvlpubs.nist.gov/nistpubs/ir/2021/NIST.IR.8312.pdf", "version_or_date": "September 2021 (doi:10.6028/NIST.IR.8312)", "source_type": "public-authority", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T09:12:00Z", "relevance": "States that a system should deliver accompanying evidence or reasons, that explanations be meaningful to the recipient, that they accurately reflect the process generating the output, and that the system declare the knowledge limits within which it operates. Grounds explanation faithfulness, audience scoping and stated knowledge limits." }, { "id": "SRC-017", "title": "Regulation (EU) 2024/1689 laying down harmonised rules on artificial intelligence (Artificial Intelligence Act)", "organization": "European Parliament and Council of the European Union", "url": "https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=OJ:L_202401689", "version_or_date": "Adopted 13 June 2024; OJ L, 2024/1689, 12.7.2024", "source_type": "legislation", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T09:14:00Z", "relevance": "Article 86 gives an affected person the right to obtain from the deployer clear and meaningful explanations of the role of the AI system in the decision-making procedure and the main elements of the decision taken; Article 13 requires clear and comprehensible information on purpose, capabilities and limitations. Grounds bounded explanation renditions and audience scoping." }, { "id": "SRC-018", "title": "CiTO, the Citation Typing Ontology", "organization": "SPAR Ontologies (Semantic Publishing and Referencing)", "url": "https://sparontologies.github.io/cito/current/cito.html", "version_or_date": "Version 2.8.2, modified 2026-06-22 (namespace http://purl.org/spar/cito/)", "source_type": "ontology", "primary_source": true, "authority_tier": 2, "accessed_at": "2026-09-02T09:16:00Z", "relevance": "Provides governed sub-properties of cites for characterising citation intent, including citesAsEvidence, citesAsAuthority, supports, obtainsSupportFrom, disagreesWith, refutes and includesQuotationFrom. Grounds citation typing and citation stance in evidence bindings." }, { "id": "SRC-019", "title": "Goal Structuring Notation Community Standard (Version 3)", "organization": "Safety-Critical Systems Club (SCSC), Assurance Case Working Group", "url": "https://scsc.uk/resources/citation_r1386.html", "version_or_date": "Version 3, published 2021-05-04 (SCSC document r1386)", "source_type": "standard", "primary_source": true, "authority_tier": 2, "accessed_at": "2026-09-02T09:20:00Z", "relevance": "Defines Goal, Solution (reference to evidence), Context, Assumption and Justification as distinct argument elements, plus a dialectic extension with Challenge and Defeated. Grounds the separation of assumption from supported premise, of justification from claim, and of recorded challenge from decision outcome." }, { "id": "SRC-020", "title": "ACIP GRADE Handbook, Chapter 7: GRADE Criteria Determining Certainty of Evidence", "organization": "Centers for Disease Control and Prevention (CDC), U.S. Department of Health and Human Services", "url": "https://www.cdc.gov/acip-grade-handbook/hcp/chapter-7-grade-criteria-determining-certainty-of-evidence/index.html", "version_or_date": "Last updated 2024-04-22", "source_type": "public-authority", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T09:24:00Z", "relevance": "Sets four certainty levels (high, moderate, low, very low) defined as degrees of confidence that the true effect lies close to the estimate, with five downgrading domains (risk of bias, inconsistency, indirectness, imprecision, publication bias) and three upgrading criteria. Grounds the distinction between an evidence body's certainty grade and a decision-local appraisal." }, { "id": "SRC-021", "title": "Consolidated version of the Treaty on the Functioning of the European Union, Article 296", "organization": "European Union", "url": "https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:12016E296", "version_or_date": "Consolidated version, OJ C 202, 7.6.2016", "source_type": "legislation", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T09:28:00Z", "relevance": "Requires that legal acts state the reasons on which they are based and refer to any proposals, initiatives, recommendations, requests or opinions required by the Treaties. Establishes the stated reason as a governed, citable object attached to a decision." }, { "id": "SRC-022", "title": "OMB Circular A-4, Regulatory Analysis", "organization": "Office of Management and Budget, Executive Office of the President of the United States", "url": "https://obamawhitehouse.archives.gov/omb/circulars_a004_a-4/", "version_or_date": "17 September 2003 (archived copy)", "source_type": "public-authority", "primary_source": true, "authority_tier": 2, "accessed_at": "2026-09-02T09:32:00Z", "relevance": "Requires a statement of need, examination of alternative approaches, sensitivity analysis showing how results vary with plausible changes in main assumptions and input data, explicit statement of assumptions used, and a report a qualified third party can follow to see how estimates and conclusions were reached. Grounds assumptions, trade-offs, sensitivity and reproducibility." }, { "id": "SRC-023", "title": "Web Annotation Data Model", "organization": "World Wide Web Consortium (W3C)", "url": "https://www.w3.org/TR/annotation-model/", "version_or_date": "W3C Recommendation, 23 February 2017 (REC-annotation-model-20170223)", "source_type": "standard", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T09:35:00Z", "relevance": "Defines body/target annotation structure, motivation values, SpecificResource with source, selector, state and purpose, selector types (TextQuoteSelector, TextPositionSelector, FragmentSelector and others) and provenance properties creator, created, generator and generated. Grounds precise locators in citation bindings without copying cited content." }, { "id": "SRC-024", "title": "ISO 15489-1:2016 Information and documentation — Records management — Part 1: Concepts and principles", "organization": "ISO/TC 46/SC 11 Archives/records management", "url": "https://committee.iso.org/sites/tc46sc11/home/projects/published/iso-15489-records-management.html", "version_or_date": "Second edition, 2016", "source_type": "standard", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T08:35:00Z", "relevance": "Establishes records characteristics (authenticity, reliability, integrity, usability), metadata for records, assigned responsibilities, records controls, appraisal and disposition — the normative basis for governance, custody and disposition findings." }, { "id": "SRC-025", "title": "Universal Electronic Records Management (ERM) Requirements", "organization": "National Archives and Records Administration (NARA), United States", "url": "https://www.archives.gov/records-mgmt/policy/universalermrequirements", "version_or_date": "Version 3, June 2023", "source_type": "public-authority", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T08:38:00Z", "relevance": "Organises electronic records requirements into Capture, Maintenance and Use, Disposal, Transfer, Metadata and Reporting, separating program from system requirements and mandatory from preferred — the basis for retention, disposition, transfer and governance metadata expectations." }, { "id": "SRC-026", "title": "Verifiable Credentials Data Model v2.0", "organization": "World Wide Web Consortium (W3C)", "url": "https://www.w3.org/TR/vc-data-model-2.0/", "version_or_date": "W3C Recommendation, 15 May 2025", "source_type": "standard", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T08:39:00Z", "relevance": "Defines issuer as the asserting authority, validFrom/validUntil validity windows, credentialStatus for revocation and suspension, and termsOfUse restrictions — the alignment basis for authority attribution, validity windows and suspended/revoked status semantics." }, { "id": "SRC-027", "title": "NIST Special Publication 800-53 Revision 5: Security and Privacy Controls for Information Systems and Organizations", "organization": "National Institute of Standards and Technology (NIST), United States", "url": "https://csrc.nist.gov/pubs/sp/800/53/r5/upd1/final", "version_or_date": "Rev. 5, September 2020 (updates through December 2020; control catalog release 5.2.0, 27 August 2025)", "source_type": "standard", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T08:40:00Z", "relevance": "Control families for Audit and Accountability (audit records, non-repudiation), Access Control (access enforcement) and System and Information Integrity (information retention and disposal) — the basis for treating audit trails and enforcement as externally owned while this model holds references and classifications." }, { "id": "SRC-028", "title": "RFC 3339: Date and Time on the Internet: Timestamps", "organization": "Internet Engineering Task Force (IETF)", "url": "https://www.rfc-editor.org/rfc/rfc3339", "version_or_date": "Proposed Standard, July 2002 (updated by RFC 9557)", "source_type": "standard", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T08:41:00Z", "relevance": "Requires an explicit time-offset of 'Z' or +/-HH:MM, defines the seconds field including leap seconds, and reserves '-00:00' for an unknown local offset — the normative basis for the timestamp rule and for separating event, observation and ingestion times." }, { "id": "SRC-029", "title": "RFC 2026 (BCP 9): The Internet Standards Process -- Revision 3", "organization": "Internet Engineering Task Force (IETF)", "url": "https://www.rfc-editor.org/rfc/rfc2026", "version_or_date": "Best Current Practice, October 1996", "source_type": "standard", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T08:42:00Z", "relevance": "Documents maturity states, revision of a specification through the full process, retirement to Historic status, and a tiered conflict-resolution and appeals procedure with a two-month submission limit — the basis for status transitions, supersession and appeal/reconsideration windows." }, { "id": "SRC-030", "title": "Regulation (EC) No 1049/2001 regarding public access to European Parliament, Council and Commission documents", "organization": "European Parliament and Council of the European Union", "url": "https://eur-lex.europa.eu/eli/reg/2001/1049/oj", "version_or_date": "30 May 2001", "source_type": "legislation", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T08:43:00Z", "relevance": "Article 4 exemption grounds including protection of an institution's decision-making process, Article 4(6) partial access by releasing non-exempt parts, Article 9 sensitive-document handling with originator consent, Article 11 public registers of documents, and the duty to give written reasons for refusal — a jurisdictional profile for access classification and disclosure exceptions." }, { "id": "SRC-031", "title": "ELI ontology (European Legislation Identifier)", "organization": "Publications Office of the European Union", "url": "https://data.europa.eu/eli/ontology", "version_or_date": "Version 1.5", "source_type": "ontology", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T08:44:00Z", "relevance": "Defines in_force, consolidates/consolidated_by, changes/changed_by, repeals/repealed_by, responsibility_of_agent and passed_by over FRBR-style legal resources and expressions — the alignment basis for effective/superseded/repealed states, consolidation and authority attribution." }, { "id": "SRC-032", "title": "Council Decision of 1 December 2009 adopting the Council's Rules of Procedure (2009/937/EU)", "organization": "Council of the European Union", "url": "https://eur-lex.europa.eu/eli/dec/2009/937/oj", "version_or_date": "1 December 2009", "source_type": "legislation", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T08:45:00Z", "relevance": "Article 11 requires the presence of a majority of members entitled to vote and permits a member to act on behalf of not more than one other member; Article 12 provides written and silence procedures; Article 13 governs minutes recording decisions and statements — a concrete primary basis for quorum attestation, proxy representation, decision rule and participation mode." }, { "id": "SRC-033", "title": "Regulation (EU) 2016/679 (General Data Protection Regulation), consolidated text", "organization": "European Parliament and Council of the European Union", "url": "https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A02016R0679-20160504", "version_or_date": "27 April 2016, consolidated 4 May 2016", "source_type": "legislation", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-09-02T08:46:00Z", "relevance": "Article 5(1)(e) storage limitation with an archiving derogation, Article 17 right to erasure and the Article 17(3) exceptions for legal obligations, public-interest tasks, archiving and legal claims, and Article 30 records of processing — the profile that constrains when erasure overrides tombstone-first deletion." } ], "structure": { "bundles": [ { "id": "dr-core-framing", "name": "Decision framing and identity", "description": "How a specific decision is identified, typed against a jurisdictional profile, and framed by its question, objectives, context, constraints and the rules it relies upon.", "rationale": "DMN's Decision element carries question and allowedAnswers, its objectives and impacted indicators, and its information and knowledge requirements; ISO/IEC/IEEE 42010 relates a decision to the concerns it addresses and the elements it affects; appraisal guidance requires a case for change, SMART objectives and critical success factors before options are considered. Framing must therefore be modelled before alternatives, and it is the part of the decision that determines what a complete record looks like.", "source_refs": [ "SRC-001", "SRC-003", "SRC-006", "SRC-008" ], "layers": [ { "id": "dr-core-identity-layer", "name": "Decision identity and classification", "description": "Stable identification of the decision content, discrimination between its expressions, and its classification against a decision type and jurisdictional profile.", "source_refs": [ "SRC-001", "SRC-005", "SRC-012" ], "findings": [ { "id": "dr-core-identity", "name": "Decision identity and version discriminator", "description": "How one governed decision is stably identified, how successive expressions of the same decision are discriminated, how the identity of the decision content is kept distinct from the identity of the record that carries it, and how supersession and alternate jurisdictional identifiers are recorded.", "source_refs": [ "SRC-001", "SRC-012", "SRC-003" ], "questions": [ { "id": "dr-core-q-id-master", "text": "Which authoritative master-system identifier designates this decision, and which registry mints and resolves it?", "kind": "identity", "answer_data": [ "Master identifier value", "Issuing system of record and identifier scheme name", "Resolution endpoint or registry reference" ] }, { "id": "dr-core-q-id-version", "text": "What version discriminator distinguishes this expression of the decision from earlier or later expressions of the same decision?", "kind": "identity", "answer_data": [ "Version discriminator value", "Rule by which the discriminator increases", "Reference to the immediately preceding expression" ] }, { "id": "dr-core-q-id-container", "text": "Is the identifier scoped to the decision content itself or to the record that carries it, and how is that distinction represented?", "kind": "composition", "answer_data": [ "Content identifier", "Carrying-record reference and its own identifier", "Statement of which identifier is authoritative for citation" ] }, { "id": "dr-core-q-id-supersession", "text": "Which prior decision, if any, does this decision supersede, amend, reaffirm or reverse, and by which identifier?", "kind": "relationship", "answer_data": [ "Prior decision identifier", "Supersession relation code", "Scope of the supersession, whole or partial" ] }, { "id": "dr-core-q-id-alternate", "text": "Which external or jurisdictional identifiers are carried as labelled alternate identifiers for this decision?", "kind": "interoperability", "answer_data": [ "Alternate identifier value", "Scheme name and issuing authority", "Statement that the alternate identifier is non-authoritative" ] } ], "data_elements": [ { "id": "dr-core-de-decision-id", "name": "Decision identifier", "description": "The authoritative identifier of the governed decision content, minted according to the identity priority order.", "value_kind": "identifier", "cardinality": "1", "required": true, "source_refs": [ "SRC-001", "SRC-012" ] }, { "id": "dr-core-de-version-discriminator", "name": "Version discriminator", "description": "Label or ordinal distinguishing this expression of the decision content from other expressions of the same decision; never a bare date.", "value_kind": "text", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-012" ] }, { "id": "dr-core-de-carrying-record-ref", "name": "Carrying record reference", "description": "Reference to the document or record container in which this decision content is expressed, held for navigation only.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-012" ] }, { "id": "dr-core-de-supersedes-ref", "name": "Supersedes reference", "description": "Identifier of a prior decision that this decision supersedes, amends, reaffirms or reverses, with the relation code.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-003" ] }, { "id": "dr-core-de-alternate-identifier", "name": "Alternate identifier", "description": "Labelled non-authoritative identifier from an external scheme, such as a case, docket or published-citation reference.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-012" ] } ], "artifacts": [], "inline_only_rationale": "Identity is reference data carried directly on the decision record: identifier values, a version discriminator, typed links to superseded decisions and labelled alternate identifiers. No separate governed artefact is produced, and any rendered identifier block belongs to the carrying record model rather than to decision content." }, { "id": "dr-core-typing", "name": "Decision type, jurisdictional profile and required content", "description": "Classification of the decision against a controlled decision-type scheme, binding of the jurisdictional or institutional profile that governs its required content, and declaration of the content elements that profile makes mandatory.", "source_refs": [ "SRC-005", "SRC-009", "SRC-010", "SRC-002" ], "questions": [ { "id": "dr-core-q-type-code", "text": "Which decision type does this record instantiate, and under which controlled classification scheme is that type published?", "kind": "classification", "answer_data": [ "Decision type code", "Classification scheme identifier and version", "Definition text of the type" ] }, { "id": "dr-core-q-type-profile", "text": "Which jurisdiction, regulatory regime or institutional profile governs the content this decision must carry?", "kind": "requirement", "answer_data": [ "Profile identifier", "Governing instrument reference", "Territorial or organisational applicability of the profile" ] }, { "id": "dr-core-q-type-automated", "text": "Does the decision produce legal or similarly significant effects for an identified natural person, and is it based solely on automated processing?", "kind": "requirement", "answer_data": [ "Legal-effect flag", "Solely-automated flag", "Applicable safeguard obligations recorded as profile requirements" ] }, { "id": "dr-core-q-type-mandatory-elements", "text": "Which content elements must be present for a decision of this type and profile to be treated as complete?", "kind": "validation", "answer_data": [ "List of profile-required content element names", "Completeness rule per element", "Consequence of absence, blocking or advisory" ] }, { "id": "dr-core-q-type-not-instrument", "text": "How is this decision type distinguished from a legal judgment or from an executable business rule covering the same subject matter?", "kind": "definition", "answer_data": [ "Distinguishing criteria text", "Reference to the legal instrument model where one exists", "Reference to the executable rule or decision service where one exists" ] } ], "data_elements": [ { "id": "dr-core-de-decision-type", "name": "Decision type code", "description": "Controlled code classifying the decision, drawn from a scheme published by the owning package.", "value_kind": "code", "cardinality": "1", "required": true, "source_refs": [ "SRC-005", "SRC-010" ] }, { "id": "dr-core-de-profile-binding", "name": "Jurisdictional profile binding", "description": "Identifier of the jurisdictional, regulatory or institutional profile whose content requirements apply to this decision.", "value_kind": "code", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-005", "SRC-009" ] }, { "id": "dr-core-de-automated-flag", "name": "Solely automated decision flag", "description": "Whether the outcome was produced solely by automated processing, and whether it produces legal or similarly significant effects for a natural person.", "value_kind": "boolean", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-005" ] }, { "id": "dr-core-de-required-elements", "name": "Profile-required content elements", "description": "Names of the content elements the bound profile makes mandatory, used for completeness validation.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-010", "SRC-009" ] } ], "artifacts": [], "inline_only_rationale": "Typing and profile binding are code values plus a derived list of required element names. They are structured reference data evaluated against published code lists held by the profile custodian; producing a separate artefact would duplicate the code-list registry rather than add governed content." } ] }, { "id": "dr-core-problem-layer", "name": "Decision question, objectives and context", "description": "What the decision is being asked to settle, what it is intended to achieve, and the context, dependencies and affected subjects within which it is taken.", "source_refs": [ "SRC-001", "SRC-006", "SRC-008", "SRC-003" ], "findings": [ { "id": "dr-core-question", "name": "Decision question, triggering issue and admissible answer space", "description": "The single question the decision answers, the admissible answers to that question, the problem or issue that made the decision necessary, and whether the question decomposes into required sub-decisions.", "source_refs": [ "SRC-001", "SRC-006", "SRC-008" ], "questions": [ { "id": "dr-core-q-question-text", "text": "What single question does this decision answer, stated so that a recorded answer can be checked against it?", "kind": "definition", "answer_data": [ "Decision question text", "Natural language and locale of the statement", "Test for whether a candidate answer responds to the question" ] }, { "id": "dr-core-q-question-answers", "text": "What is the admissible answer space for that question, and is it an enumerated set or an open expression?", "kind": "constraint", "answer_data": [ "Allowed answer statement or enumeration", "Whether the space is closed or open", "Datatype or unit of an admissible answer" ] }, { "id": "dr-core-q-question-trigger", "text": "What problem, need or issue made this decision necessary, and which upstream item raised it?", "kind": "relationship", "answer_data": [ "Problem statement text", "Triggering issue or change-request reference", "Rationale for intervention, such as a stated market failure or equity concern" ] }, { "id": "dr-core-q-question-decomposition", "text": "Is this an atomic decision, or does it decompose into sub-decisions that must be settled for the answer to hold?", "kind": "composition", "answer_data": [ "Atomic or composite indicator", "References to required sub-decisions", "Dependency ordering among sub-decisions" ] } ], "data_elements": [ { "id": "dr-core-de-decision-question", "name": "Decision question", "description": "The question the decision answers, expressed as a single interrogative statement.", "value_kind": "text", "cardinality": "1", "required": true, "source_refs": [ "SRC-001" ] }, { "id": "dr-core-de-allowed-answers", "name": "Admissible answer space", "description": "Statement or enumeration of the answers admissible for the decision question.", "value_kind": "text", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-001" ] }, { "id": "dr-core-de-problem-statement", "name": "Problem statement", "description": "The need, issue or case for change that makes the decision necessary.", "value_kind": "text", "cardinality": "1", "required": true, "source_refs": [ "SRC-006", "SRC-008" ] }, { "id": "dr-core-de-triggering-issue-ref", "name": "Triggering issue reference", "description": "Reference to the change request, incident or request that the decision is taken to resolve.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-006" ] }, { "id": "dr-core-de-sub-decision-ref", "name": "Sub-decision reference", "description": "Reference to a decision whose outcome this decision requires.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-001" ] } ], "artifacts": [], "inline_only_rationale": "The question, its admissible answers and the problem statement are short authored text fields plus typed references. DMN carries them as simple string children of the Decision element, and holding them inline keeps the answer directly comparable with the recorded outcome statement without a separate artefact to keep in step." }, { "id": "dr-core-objectives", "name": "Objectives, success factors and expected effects", "description": "The objectives the decision is intended to achieve, the critical success factors any acceptable option must satisfy, the indicators expected to be affected and the horizon over which effects are expected.", "source_refs": [ "SRC-001", "SRC-006", "SRC-008" ], "questions": [ { "id": "dr-core-q-obj-statement", "text": "Which objectives is the decision intended to achieve, expressed specifically enough to be measurable?", "kind": "requirement", "answer_data": [ "Objective statement", "Beneficiary, quantity and location where applicable", "Whether the objective is outcome-based or output-based" ] }, { "id": "dr-core-q-obj-csf", "text": "Which critical success factors must any acceptable option possess for the objectives to be delivered?", "kind": "constraint", "answer_data": [ "Critical success factor statement", "Category, such as strategic fit, value for money, capacity, affordability or achievability", "Whether the factor is pass or fail at longlist stage" ] }, { "id": "dr-core-q-obj-indicator", "text": "Which performance indicators or outcome measures are expected to be affected, and in which direction?", "kind": "measurement", "answer_data": [ "Indicator reference", "Expected direction of movement", "Baseline value and unit where stated" ] }, { "id": "dr-core-q-obj-horizon", "text": "By when are the desired effects expected to be realised, and over what appraisal period are they assessed?", "kind": "temporal", "answer_data": [ "Expected realisation date or duration", "Appraisal or assessment period", "Time basis, such as from effect onset or from implementation" ] } ], "data_elements": [ { "id": "dr-core-de-objective", "name": "Objective", "description": "A stated objective the decision is intended to achieve, expressed in specific and measurable terms.", "value_kind": "collection", "cardinality": "1..n", "required": true, "source_refs": [ "SRC-006", "SRC-001" ] }, { "id": "dr-core-de-critical-success-factor", "name": "Critical success factor", "description": "An attribute any acceptable option must possess for the objectives to be delivered.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-006" ] }, { "id": "dr-core-de-impacted-indicator-ref", "name": "Impacted indicator reference", "description": "Reference to a performance indicator the decision is expected to affect, with expected direction.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-001" ] }, { "id": "dr-core-de-effect-horizon", "name": "Effect horizon", "description": "Period over which the desired effects are expected to be realised and assessed.", "value_kind": "duration", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-006" ] } ], "artifacts": [], "inline_only_rationale": "Objectives, success factors and indicator references are short structured statements bound directly to the decision. DMN models supportedObjective and impactedPerformanceIndicator as element references rather than documents, and the indicator definitions themselves are governed by a measurement model, so no artefact is produced here." }, { "id": "dr-core-context-scope", "name": "Decision context, dependencies and affected subjects", "description": "The subjects, populations, assets or system elements the decision affects, the extent over which it applies, the input data and upstream decisions it depends on, the concerns it addresses and the assumptions declared about its context.", "source_refs": [ "SRC-001", "SRC-003", "SRC-005", "SRC-009" ], "questions": [ { "id": "dr-core-q-ctx-subjects", "text": "Which subjects, populations, assets or system elements are affected by this decision?", "kind": "relationship", "answer_data": [ "Affected subject reference and subject class", "Nature of the effect on each subject", "Whether any affected subject is an identified natural person" ] }, { "id": "dr-core-q-ctx-extent", "text": "Over what geographic, organisational or system extent does the decision apply?", "kind": "spatial", "answer_data": [ "Named extent or organisational unit reference", "Geometry or boundary description where a spatial extent applies", "Explicit statement where the extent is unbounded" ] }, { "id": "dr-core-q-ctx-dependencies", "text": "Which input data, upstream decisions and knowledge sources does this decision depend on?", "kind": "composition", "answer_data": [ "Required input data reference", "Upstream decision reference", "Knowledge source reference and its location where published" ] }, { "id": "dr-core-q-ctx-concerns", "text": "Which concerns or stakeholder interests does the decision pertain to, and which of them does it leave unaddressed?", "kind": "relationship", "answer_data": [ "Addressed concern statement or reference", "Concern owner class", "Concerns explicitly deferred" ] }, { "id": "dr-core-q-ctx-assumptions", "text": "Which assumptions about the context are declared, and which of them are known to be unverified?", "kind": "quality", "answer_data": [ "Assumption statement", "Verification status of the assumption", "Effect on the decision if the assumption fails" ] } ], "data_elements": [ { "id": "dr-core-de-affected-subject-ref", "name": "Affected subject reference", "description": "Reference to a subject, population, asset or system element affected by the decision, with the nature of the effect.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-003", "SRC-005" ] }, { "id": "dr-core-de-applicability-extent", "name": "Applicability extent", "description": "Geographic, organisational or system extent over which the decision applies.", "value_kind": "text", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-009" ] }, { "id": "dr-core-de-required-input-ref", "name": "Required input reference", "description": "Reference to input data or a knowledge source the decision depends on.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-001" ] }, { "id": "dr-core-de-upstream-decision-ref", "name": "Upstream decision reference", "description": "Reference to another decision whose outcome this decision requires as an input.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-001" ] }, { "id": "dr-core-de-declared-assumption", "name": "Declared assumption", "description": "A contextual assumption the decision relies on, with its verification status.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-008" ] }, { "id": "dr-core-de-addressed-concern", "name": "Addressed concern", "description": "A concern or stakeholder interest the decision pertains to.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-003" ] } ], "artifacts": [ { "id": "dr-core-art-drg", "name": "Decision requirements graph extract", "description": "Format-neutral dependency graph naming the input data, upstream decisions and knowledge sources this decision requires, with typed edges for each dependency. It carries dependency references only: authority-requirement semantics, executable logic and any runtime resolution of the graph stay with the authority model and the decision service.", "media_or_form": [ "node and edge dependency graph", "structured dependency list", "rendered dependency diagram" ], "serial": false, "identity_strategy": "Bound to the decision's authoritative master-system identifier plus a fixed graph-extract discriminator; nodes reuse the identifiers of the referenced input, decision and knowledge-source records rather than minting new ones.", "source_refs": [ "SRC-001", "SRC-003" ] } ], "inline_only_rationale": null } ] }, { "id": "dr-core-constraint-layer", "name": "Constraints and applied decision rules", "description": "The binding conditions any outcome must satisfy and the declarative rules the decision relies on to map inputs to admissible outcomes.", "source_refs": [ "SRC-001", "SRC-006", "SRC-008" ], "findings": [ { "id": "dr-core-constraints", "name": "Binding constraints and admissibility conditions", "description": "The hard conditions any selected alternative must satisfy, their declared values and units, the separation of binding constraints from soft preferences, and the treatment of an alternative that violates one.", "source_refs": [ "SRC-006", "SRC-008", "SRC-004" ], "questions": [ { "id": "dr-core-q-con-hard", "text": "Which hard constraints must any selected alternative satisfy, and which external requirement does each constraint cite?", "kind": "constraint", "answer_data": [ "Constraint statement", "Cited requirement or governing instrument reference", "Whether the constraint is legal, technical, financial or temporal" ] }, { "id": "dr-core-q-con-values", "text": "Which constraints express a budgetary, temporal or capacity limit, and what value and unit does each carry?", "kind": "measurement", "answer_data": [ "Limit value with unit", "Comparison operator against the measured quantity", "Price base or time basis where relevant" ] }, { "id": "dr-core-q-con-soft", "text": "Which stated conditions are soft preferences rather than binding constraints, and how is that distinction recorded?", "kind": "classification", "answer_data": [ "Binding or advisory flag per condition", "Basis for the classification", "Where a soft preference is instead handled as a weighted criterion" ] }, { "id": "dr-core-q-con-violation", "text": "What happens to an alternative that violates a binding constraint, and can a waiver admit it despite the violation?", "kind": "exception", "answer_data": [ "Disposition of the violating alternative", "Waiver condition text where a waiver is possible", "Reference to the instrument that would grant the waiver" ] } ], "data_elements": [ { "id": "dr-core-de-constraint", "name": "Constraint", "description": "A stated condition on the decision, with a flag marking whether it is binding or advisory.", "value_kind": "collection", "cardinality": "1..n", "required": true, "source_refs": [ "SRC-006", "SRC-008" ] }, { "id": "dr-core-de-constraint-value", "name": "Constraint limit value", "description": "Numeric limit and unit for a budgetary, temporal or capacity constraint.", "value_kind": "quantity", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-006" ] }, { "id": "dr-core-de-constraint-source-ref", "name": "Constraint source reference", "description": "Reference to the requirement, instrument or standard that the constraint cites.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-004" ] }, { "id": "dr-core-de-waiver-condition", "name": "Waiver condition", "description": "Condition under which an alternative violating a binding constraint could still be admitted.", "value_kind": "text", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-008" ] } ], "artifacts": [], "inline_only_rationale": "Constraints are short structured statements with a binding flag, an optional quantified limit and a citation. They are read alongside the criterion set during screening, and holding them inline avoids a second place where a limit could drift out of step with the criterion that measures it." }, { "id": "dr-core-decision-rules", "name": "Applied decision rules and conflict-resolution policy", "description": "The declarative rules the decision relies on to map stated inputs to admissible outcomes, the declared policy for overlapping rules, the versioned reference to any external rule source, and whether the outcome was rule-determined or discretionary. Statement-level only: executing the rules is the work of a decision service outside this model.", "source_refs": [ "SRC-001", "SRC-002", "SRC-005" ], "questions": [ { "id": "dr-core-q-rule-set", "text": "Which decision rules map the stated inputs to admissible outcomes for this decision?", "kind": "requirement", "answer_data": [ "Rule condition entries per input", "Rule conclusion or output entry", "Ordering or priority among rules where declared" ] }, { "id": "dr-core-q-rule-hit", "text": "When several rules match, which conflict-resolution or hit policy is declared, and does it yield a single result or several?", "kind": "constraint", "answer_data": [ "Hit policy code", "Single-result or multiple-result indicator", "Aggregation applied to multiple results where declared" ] }, { "id": "dr-core-q-rule-source", "text": "Which external executable rule, policy or model is cited as the source of the applied rules, and at which version?", "kind": "interoperability", "answer_data": [ "External rule source reference", "Source version identifier", "Statement that execution and enforcement belong to the referenced service" ] }, { "id": "dr-core-q-rule-discretion", "text": "How is a discretionary outcome, where no rule determines the answer, distinguished from a rule-determined one?", "kind": "classification", "answer_data": [ "Discretion flag", "Extent of discretion exercised", "Which rules constrained the discretionary space" ] } ], "data_elements": [ { "id": "dr-core-de-applied-rule", "name": "Applied decision rule", "description": "A declarative rule relied upon by the decision, expressed as input conditions and an output entry.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-001" ] }, { "id": "dr-core-de-hit-policy", "name": "Hit policy code", "description": "Declared policy for resolving overlapping rules, aligned to the DMN hit policy vocabulary.", "value_kind": "code", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-001" ] }, { "id": "dr-core-de-rule-source-ref", "name": "Rule source reference", "description": "Versioned reference to the external executable rule, policy or knowledge model from which the applied rules derive.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-002", "SRC-005" ] }, { "id": "dr-core-de-discretion-flag", "name": "Discretion flag", "description": "Whether the outcome was determined by the applied rules or reached by exercised discretion.", "value_kind": "boolean", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-005" ] } ], "artifacts": [ { "id": "dr-core-art-rule-table", "name": "Applied decision rule table", "description": "Tabular statement of the input conditions, output entries and declared hit policy the decision relied upon, with an annotation column for the citation of each rule. It is a statement of the rules applied, not a deployed executable artefact: evaluation, enforcement and execution logging remain with the referenced decision service.", "media_or_form": [ "decision table with rules as rows", "structured condition and conclusion list", "DMN-aligned interchange form" ], "serial": false, "identity_strategy": "Bound to the decision's authoritative master-system identifier plus a rule-table discriminator; the cited external rule source keeps its own governed identifier and version, which the table records rather than replaces.", "source_refs": [ "SRC-001", "SRC-002" ] } ], "inline_only_rationale": null } ] } ] }, { "id": "dr-core-options", "name": "Alternatives, screening and evaluation", "description": "The candidate alternatives placed under consideration including any required baseline, the screening that determines eligibility, the criteria and weights used for comparison, and the recorded evaluation results.", "rationale": "Appraisal and decision-analysis guidance treats the alternative set as a governed object in its own right: HM Treasury requires a longlist narrowed to a shortlist with business-as-usual always carried forward, NEPA practice required inclusion of the no action alternative and recorded reasons for eliminating others, and NASA requires screening against mandatory criteria before scored evaluation. Criteria, weights and results are therefore modelled separately from framing and from the outcome, so that a comparison can be inspected and re-run without touching the framing or the recorded answer.", "source_refs": [ "SRC-006", "SRC-007", "SRC-008", "SRC-009" ], "layers": [ { "id": "dr-core-alternatives-layer", "name": "Candidate alternatives and eligibility", "description": "Registration of the alternatives under consideration, including any profile-required baseline, and the screening that admits or excludes each from detailed evaluation.", "source_refs": [ "SRC-006", "SRC-008", "SRC-009" ], "findings": [ { "id": "dr-core-alternative-set", "name": "Candidate alternative set and mandatory baseline", "description": "The alternatives placed under consideration, how each is identified and described, whether a no-action or business-as-usual baseline is present as the profile requires, and how variants of one alternative are distinguished from genuinely distinct alternatives.", "source_refs": [ "SRC-006", "SRC-009", "SRC-008" ], "questions": [ { "id": "dr-core-q-alt-register", "text": "Which candidate alternatives were placed under consideration, and how is each identified within this decision?", "kind": "identity", "answer_data": [ "Alternative identifier and registration ordinal", "Short name of the alternative", "Stage at which it entered consideration, longlist or shortlist" ] }, { "id": "dr-core-q-alt-baseline", "text": "Is a no-action, business-as-usual or do-minimum baseline included, and if not, on what stated basis is it absent?", "kind": "requirement", "answer_data": [ "Baseline alternative reference", "Baseline kind, no action, business as usual or do minimum", "Recorded justification where the bound profile allows its absence" ] }, { "id": "dr-core-q-alt-variant", "text": "How are variants of the same alternative distinguished from genuinely distinct alternatives?", "kind": "classification", "answer_data": [ "Variant relation to a parent alternative", "Dimension on which the variant differs, such as scope, solution, delivery, implementation or funding", "Whether the variant is scored independently" ] }, { "id": "dr-core-q-alt-content", "text": "What does each alternative commit to doing, described in terms sufficient for it to be evaluated against the declared criteria?", "kind": "definition", "answer_data": [ "Alternative description text", "Distinguishing features relative to the baseline", "Assumptions specific to that alternative" ] } ], "data_elements": [ { "id": "dr-core-de-alternative", "name": "Alternative", "description": "A registered candidate alternative with its identifier, name, description and registration ordinal.", "value_kind": "collection", "cardinality": "1..n", "required": true, "source_refs": [ "SRC-006", "SRC-008" ] }, { "id": "dr-core-de-alternative-kind", "name": "Alternative kind code", "description": "Code marking an alternative as the baseline, a do-minimum, a preferred way forward, a more ambitious or less ambitious variant, or an ordinary candidate.", "value_kind": "code", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-006" ] }, { "id": "dr-core-de-baseline-ref", "name": "Baseline alternative reference", "description": "Reference to the alternative that serves as the no-action or business-as-usual benchmark for comparison.", "value_kind": "reference", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-009", "SRC-006" ] }, { "id": "dr-core-de-alternative-variant-of", "name": "Variant-of reference", "description": "Reference from a variant to the alternative it varies, with the dimension of variation.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-006" ] } ], "artifacts": [], "inline_only_rationale": "Alternatives are append-only structured records held on the decision: an identifier, a kind code, a description and a variant link. They must stay directly addressable from evaluation results and from the outcome, so extracting them into a separate artefact would add an indirection with no governance benefit." }, { "id": "dr-core-screening", "name": "Eligibility screening and recorded exclusion", "description": "The screening or mandatory criteria that decide whether an alternative proceeds to detailed evaluation, the recorded reason for each exclusion, the stage at which it occurred, and the conditions under which an excluded alternative could be reinstated.", "source_refs": [ "SRC-009", "SRC-006", "SRC-008" ], "questions": [ { "id": "dr-core-q-scr-criteria", "text": "Which screening or mandatory criteria decide whether an alternative is eligible for detailed evaluation?", "kind": "validation", "answer_data": [ "Screening criterion statement", "Pass or fail threshold", "Whether failure is absolute or reviewable" ] }, { "id": "dr-core-q-scr-excluded", "text": "Which alternatives were eliminated from detailed study, and what reason is recorded against each?", "kind": "exception", "answer_data": [ "Excluded alternative reference", "Recorded exclusion reason text", "Screening criterion or constraint that the alternative failed" ] }, { "id": "dr-core-q-scr-stage", "text": "Which recorded screening stage does each exclusion cite, and was the exclusion by filter rating, mandatory-criterion failure or constraint violation?", "kind": "process", "answer_data": [ "Screening stage identifier", "Exclusion mechanism code", "Rating assigned where a filter rating scheme was used" ] }, { "id": "dr-core-q-scr-reinstate", "text": "Can an excluded alternative be reinstated, and what would have to change for that to happen?", "kind": "constraint", "answer_data": [ "Reinstatement condition text", "Whether reinstatement requires a new evaluation round", "Whether reinstatement requires a superseding decision" ] } ], "data_elements": [ { "id": "dr-core-de-screening-criterion", "name": "Screening criterion", "description": "A mandatory or filter criterion determining eligibility of an alternative for detailed evaluation.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-008", "SRC-006" ] }, { "id": "dr-core-de-eligibility-status", "name": "Eligibility status", "description": "Per-alternative status recording whether it was admitted to detailed evaluation or excluded.", "value_kind": "code", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-009" ] }, { "id": "dr-core-de-exclusion-reason", "name": "Exclusion reason", "description": "Recorded reason why an alternative was eliminated from detailed study.", "value_kind": "text", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-009" ] }, { "id": "dr-core-de-screening-stage", "name": "Screening stage code", "description": "Code naming the stage at which the screening result was recorded, such as longlist filtering or mandatory-criterion check.", "value_kind": "code", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-006" ] }, { "id": "dr-core-de-reinstatement-condition", "name": "Reinstatement condition", "description": "Condition under which an excluded alternative could re-enter detailed evaluation.", "value_kind": "text", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-008" ] } ], "artifacts": [], "inline_only_rationale": "Screening produces a status and a short recorded reason per alternative, held directly on the alternative it concerns. The regulatory requirement is that reasons for elimination be briefly discussed, not that a separate screening document exist, and the screening activity itself belongs to the referenced decision-making activity model." } ] }, { "id": "dr-core-evaluation-layer", "name": "Criteria and evaluation results", "description": "The declared basis for comparing alternatives and the recorded results of applying it, including aggregation, ranking and sensitivity.", "source_refs": [ "SRC-007", "SRC-008", "SRC-006" ], "findings": [ { "id": "dr-core-criteria", "name": "Evaluation criteria, measures, scales and weights", "description": "The criteria against which alternatives are compared, each with its measure, scale, unit and direction of preference; the separation of mandatory pass-or-fail criteria from compensatory trade-off criteria; and the weighting scheme where one is used, including the declared weighting method.", "source_refs": [ "SRC-007", "SRC-008", "SRC-006" ], "questions": [ { "id": "dr-core-q-crit-define", "text": "Which criteria compare the alternatives, and what measure, scale and unit does each criterion use?", "kind": "measurement", "answer_data": [ "Criterion name and definition", "Measure and scale, quantitative or qualitative", "Unit and direction of preference" ] }, { "id": "dr-core-q-crit-independence", "text": "Are the declared criteria mutually independent and non-overlapping, and how was that checked?", "kind": "validation", "answer_data": [ "Independence check result per criterion pair", "Known overlaps or causal relations between criteria", "Coverage check that key performance aspects are represented" ] }, { "id": "dr-core-q-crit-weights", "text": "What weight does each criterion carry, and by which declared weighting method were the weights derived?", "kind": "decision", "answer_data": [ "Weight value per criterion", "Weighting method identifier, such as swing weighting", "Statement where no weighting is used" ] }, { "id": "dr-core-q-crit-compensatory", "text": "Which criteria are mandatory pass-or-fail and which are trade-off criteria whose shortfall can be compensated?", "kind": "classification", "answer_data": [ "Mandatory or enhancing classification per criterion", "Compensation rule where trade-off is permitted", "Link to the screening criterion where a criterion is also a filter" ] }, { "id": "dr-core-q-crit-monetised", "text": "Are any criteria monetised, and if so under which appraisal convention and price base?", "kind": "interoperability", "answer_data": [ "Monetisation flag per criterion", "Appraisal convention reference", "Currency and price base where monetised" ] } ], "data_elements": [ { "id": "dr-core-de-criterion", "name": "Criterion", "description": "A declared basis for comparing alternatives, with its definition, measure and scale.", "value_kind": "collection", "cardinality": "1..n", "required": true, "source_refs": [ "SRC-007", "SRC-008" ] }, { "id": "dr-core-de-preference-direction", "name": "Preference direction", "description": "Whether a higher or lower value on the criterion is preferred.", "value_kind": "code", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-007" ] }, { "id": "dr-core-de-criterion-weight", "name": "Criterion weight", "description": "Numeric weight assigned to a criterion under the declared weighting method.", "value_kind": "number", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-007" ] }, { "id": "dr-core-de-weighting-method", "name": "Weighting method code", "description": "Identifier of the method used to derive weights, or an explicit marker that no weighting is applied.", "value_kind": "code", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-007" ] }, { "id": "dr-core-de-compensatory-flag", "name": "Compensatory flag", "description": "Whether a shortfall on the criterion can be compensated by performance on other criteria.", "value_kind": "boolean", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-008" ] }, { "id": "dr-core-de-monetisation-basis", "name": "Monetisation basis", "description": "Appraisal convention, currency and price base where a criterion is expressed in monetary terms.", "value_kind": "text", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-006" ] } ], "artifacts": [], "inline_only_rationale": "The criterion set is a declared list of definitions, scales, directions and optional weights, addressed by identifier from every evaluation result. It is reference data whose value lies in being fixed and directly citable; the workshop or panel that derived weights is an activity governed elsewhere, and named decision-analysis methods bind as external references rather than as content held here." }, { "id": "dr-core-evaluation-results", "name": "Evaluation results, comparison and sensitivity", "description": "The recorded result of each alternative against each criterion for one evaluation round, any aggregation and ranking derived from those results, the sensitivity of the ranking to changed scores, weights or assumptions, the evidence cited by reference in support of results, and the status of contested or provisional results.", "source_refs": [ "SRC-006", "SRC-007", "SRC-008", "SRC-005" ], "questions": [ { "id": "dr-core-q-eval-scores", "text": "What result does each alternative achieve on each declared criterion, in the criterion's own measure?", "kind": "measurement", "answer_data": [ "Result value per alternative and criterion", "Unit or scale point", "Whether the result is measured, estimated or asserted" ] }, { "id": "dr-core-q-eval-ranking", "text": "What ranking or comparison follows from the recorded results, and by which aggregation was it derived?", "kind": "decision", "answer_data": [ "Aggregate value per alternative", "Ranking position per alternative", "Aggregation rule applied, including any monetised summary measure" ] }, { "id": "dr-core-q-eval-sensitivity", "text": "How sensitive is the ranking to plausible changes in scores, weights or declared assumptions?", "kind": "quality", "answer_data": [ "Sensitivity test description and the parameter varied", "Range over which the ranking is stable", "Any assumption whose failure reverses the ranking" ] }, { "id": "dr-core-q-eval-evidence", "text": "Which evidence items are cited in support of each recorded result, and what role does each citation play?", "kind": "evidence", "answer_data": [ "Evidence item reference", "Role of the citation, such as supporting a score or an assumption", "Statement that appraisal of the evidence stays with the evidence model" ] }, { "id": "dr-core-q-eval-status", "text": "Which recorded results are contested, provisional or adjusted for known bias?", "kind": "validation", "answer_data": [ "Result status code", "Nature of the contest or provisional condition", "Bias adjustment applied and its basis" ] } ], "data_elements": [ { "id": "dr-core-de-evaluation-round", "name": "Evaluation round", "description": "An append-only round of evaluation, identified by a monotonically increasing ordinal, containing results for the alternatives evaluated in it.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-007", "SRC-008" ] }, { "id": "dr-core-de-criterion-result", "name": "Criterion result", "description": "The result achieved by one alternative on one criterion within a round.", "value_kind": "number", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-007" ] }, { "id": "dr-core-de-aggregate-result", "name": "Aggregate result", "description": "Aggregated value for an alternative derived from its criterion results under the declared aggregation rule.", "value_kind": "number", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-006" ] }, { "id": "dr-core-de-ranking-position", "name": "Ranking position", "description": "Position of an alternative in the ordering produced by the round.", "value_kind": "number", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-007" ] }, { "id": "dr-core-de-sensitivity-finding", "name": "Sensitivity finding", "description": "Statement of how the ranking responds to a varied score, weight or assumption.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-007", "SRC-008" ] }, { "id": "dr-core-de-cited-evidence-ref", "name": "Cited evidence reference", "description": "Reference to an evidence item cited in support of a result or assumption, with the role of the citation.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-008", "SRC-005" ] }, { "id": "dr-core-de-result-status", "name": "Result status code", "description": "Status marking a result as settled, provisional, contested or bias-adjusted.", "value_kind": "code", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-006" ] } ], "artifacts": [ { "id": "dr-core-art-comparison-matrix", "name": "Alternative-by-criterion comparison matrix", "description": "Matrix recording each admitted alternative's result against each declared criterion for one evaluation round, with aggregation, ranking, sensitivity annotations and citation slots referencing evidence by identifier. It is a governed record of what the comparison showed; it does not execute the comparison and does not carry evidence content.", "media_or_form": [ "tabular matrix with alternatives and criteria as axes", "appraisal summary table", "structured result set" ], "serial": true, "identity_strategy": "Decision master identifier plus a zero-padded, monotonically increasing evaluation-round ordinal; the ordinal is a sequence number and never encodes a date. Superseded rounds keep their ordinal and remain resolvable.", "source_refs": [ "SRC-006", "SRC-007", "SRC-008" ] } ], "inline_only_rationale": null } ] } ] }, { "id": "dr-core-outcome", "name": "Selection, disposition and stated consequences", "description": "The alternative selected, the disposition and explicit outcome statement that answers the decision question, and the conditions, reservations and consequences the decision states as part of its own content.", "rationale": "A record-of-decision pattern requires the decision to be stated, the alternatives considered to be identified, the preferable alternative to be named, the factors balanced to be discussed, and mitigation and monitoring conditions to be declared or their absence explained. ISO/IEC/IEEE 42010 adds that making a decision can raise new concerns. The outcome is therefore modelled as governed content in its own right, distinct from the approval that authorised it and from the record that publishes it.", "source_refs": [ "SRC-010", "SRC-009", "SRC-003", "SRC-006" ], "layers": [ { "id": "dr-core-outcome-layer", "name": "Decision outcome and attached conditions", "description": "The explicit answer to the decision question with its disposition and applicability, and the conditions, unresolved matters and consequences stated as part of the decision.", "source_refs": [ "SRC-010", "SRC-003", "SRC-008" ], "findings": [ { "id": "dr-core-selection", "name": "Selected alternative, disposition and outcome statement", "description": "Which alternative is selected, the disposition the decision reaches, the explicit outcome statement that answers the decision question, the declared onset and applicability of effect, the factors expressly balanced, and any explained divergence from the evaluation ranking.", "source_refs": [ "SRC-010", "SRC-009", "SRC-006", "SRC-001" ], "questions": [ { "id": "dr-core-q-sel-alternative", "text": "Which alternative is selected, and is the disposition an approval, refusal, deferral, partial approval or a declared variant of a registered alternative?", "kind": "decision", "answer_data": [ "Selected alternative reference", "Disposition code", "Description of the variant where the selection is not a registered alternative verbatim" ] }, { "id": "dr-core-q-sel-statement", "text": "What is the explicit outcome statement, phrased as a direct answer to the recorded decision question?", "kind": "definition", "answer_data": [ "Outcome statement text", "Check that the statement falls within the admissible answer space", "Language and locale of the statement" ] }, { "id": "dr-core-q-sel-effect-time", "text": "From when does the decision take effect, and for how long or until what event does it apply?", "kind": "temporal", "answer_data": [ "Effect onset instant in RFC 3339 with seconds and offset", "Applicability end instant or terminating event", "Whether effect is immediate, deferred or conditional on an act by another party" ] }, { "id": "dr-core-q-sel-balanced", "text": "Which factors were expressly balanced in reaching this selection, including any consideration of overriding policy?", "kind": "relationship", "answer_data": [ "Balanced factor statement", "Weight or priority the factor was given in the balance", "Reference to the policy consideration where one was decisive" ] }, { "id": "dr-core-q-sel-divergence", "text": "How does the selection relate to the ranking produced by the evaluation, and is any divergence explained or a preferable alternative named separately?", "kind": "classification", "answer_data": [ "Ranking position of the selected alternative", "Divergence explanation where the top-ranked alternative was not selected", "Separately identified preferable alternative where the profile requires one" ] } ], "data_elements": [ { "id": "dr-core-de-selected-alternative-ref", "name": "Selected alternative reference", "description": "Reference to the registered alternative the decision selects, or to the alternative from which the selected variant derives.", "value_kind": "reference", "cardinality": "1", "required": true, "source_refs": [ "SRC-010" ] }, { "id": "dr-core-de-disposition", "name": "Disposition code", "description": "Controlled code naming the disposition reached, such as approved, refused, deferred, partially approved or withdrawn from consideration.", "value_kind": "code", "cardinality": "1", "required": true, "source_refs": [ "SRC-010", "SRC-005" ] }, { "id": "dr-core-de-outcome-statement", "name": "Outcome statement", "description": "Explicit statement of the decision, phrased as the answer to the recorded decision question.", "value_kind": "text", "cardinality": "1", "required": true, "source_refs": [ "SRC-010", "SRC-001" ] }, { "id": "dr-core-de-effect-onset", "name": "Effect onset", "description": "Instant from which the decision takes effect, recorded in RFC 3339 with seconds and an explicit offset or Z.", "value_kind": "timestamp", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-010" ] }, { "id": "dr-core-de-applicability-end", "name": "Applicability end", "description": "Instant or terminating event at which the decision ceases to apply.", "value_kind": "timestamp", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-010" ] }, { "id": "dr-core-de-balanced-factor", "name": "Balanced factor", "description": "A factor or policy consideration expressly balanced in reaching the selection.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-010" ] }, { "id": "dr-core-de-divergence-explanation", "name": "Divergence explanation", "description": "Recorded explanation where the selected alternative is not the one the evaluation ranked highest.", "value_kind": "text", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-006" ] }, { "id": "dr-core-de-preferable-alternative-ref", "name": "Preferable alternative reference", "description": "Separately identified alternative named as preferable on a stated dimension, where the bound profile requires such identification.", "value_kind": "reference", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-009", "SRC-010" ] } ], "artifacts": [], "inline_only_rationale": "The outcome is a small, immutable set of fields on the decision itself: a selected-alternative reference, a disposition code, the statement answering the question and the declared effect times. The published record of decision that renders these fields is a document governed by the record model, and the approval that authorised them is an activity governed elsewhere, so producing an artefact here would claim ownership of one or both." }, { "id": "dr-core-conditions", "name": "Attached conditions, unresolved matters and stated consequences", "description": "The conditions, mitigations and safeguards attached to the outcome as part of the decision itself, the matters expressly left unresolved or reserved, the consequences and accepted risks the decision states, the declared reopening or review triggers, and the new concerns or follow-on decisions the decision raises.", "source_refs": [ "SRC-010", "SRC-003", "SRC-006", "SRC-008" ], "questions": [ { "id": "dr-core-q-cond-attached", "text": "Which conditions, mitigations or safeguards are attached to the outcome as part of the decision itself?", "kind": "requirement", "answer_data": [ "Condition statement", "Whether the condition is a precondition of effect or an ongoing obligation", "Reference to the instrument that would give the condition force" ] }, { "id": "dr-core-q-cond-unresolved", "text": "Which matters are expressly left unresolved, deferred or reserved for a later decision?", "kind": "exception", "answer_data": [ "Unresolved matter statement", "Reason for deferral", "Reference to the later decision where one is already identified" ] }, { "id": "dr-core-q-cond-consequences", "text": "Which consequences does the decision state will follow, including accepted risks and residual trade-offs?", "kind": "quality", "answer_data": [ "Stated consequence text", "Accepted risk statement and the reason for accepting it", "Residual trade-off against an objective or criterion" ] }, { "id": "dr-core-q-cond-reopen", "text": "Which review or reopening trigger is declared, and what event would cause the decision to be revisited?", "kind": "event", "answer_data": [ "Reopening trigger statement", "Triggering event or threshold", "Whether reopening requires a superseding decision" ] }, { "id": "dr-core-q-cond-raised", "text": "Which new concerns or follow-on decisions does this decision raise?", "kind": "relationship", "answer_data": [ "Raised concern statement or reference", "Follow-on decision reference where identified", "Dependency direction between this decision and the follow-on" ] } ], "data_elements": [ { "id": "dr-core-de-attached-condition", "name": "Attached condition", "description": "A condition, mitigation or safeguard declared as part of the decision outcome.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-010" ] }, { "id": "dr-core-de-unresolved-matter", "name": "Unresolved matter", "description": "A matter expressly left unresolved, deferred or reserved for a later decision.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-003" ] }, { "id": "dr-core-de-accepted-risk", "name": "Accepted risk", "description": "A risk the decision states it accepts, with the reason for acceptance.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-006", "SRC-008" ] }, { "id": "dr-core-de-stated-consequence", "name": "Stated consequence", "description": "A consequence the decision states will follow from the selected outcome.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-003", "SRC-010" ] }, { "id": "dr-core-de-reopening-trigger", "name": "Reopening trigger", "description": "Declared event or threshold that would cause the decision to be reviewed or reopened.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-010" ] }, { "id": "dr-core-de-raised-concern", "name": "Raised concern", "description": "A new concern the decision raises, which may require a further decision.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-003" ] }, { "id": "dr-core-de-follow-on-decision-ref", "name": "Follow-on decision reference", "description": "Reference to a decision that this decision defers a matter to or makes necessary.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-003" ] } ], "artifacts": [], "inline_only_rationale": "Conditions, reservations and stated consequences are declarative clauses of the outcome and must travel with it in any projection. Where a decision directs a monitoring or compliance plan, that plan is an implementation artefact owned by the change and activity models; recording it here as an artefact would take ownership of execution and monitoring that this model explicitly disclaims." } ] } ] }, { "id": "dr-rsn-bundle-reasoning", "name": "Stated reasons and argument structure", "description": "The reasons offered for a decision and the argument that connects premises, inference schemes and conclusions to the recorded outcome, including contrastive reasons for not selecting alternatives and any explicitly asserted causal or counterfactual reasoning.", "rationale": "Structured-argument standards consistently separate the asserted conclusion from the inference link that supports it and from elements that conflict with it, while the EU duty to state reasons makes the stated reason itself a governed object attached to a legal act. Keeping the offered reason and the argument connecting it to the outcome as separate, independently citable structures prevents a narrative summary from being mistaken for a checked inference.", "source_refs": [ "SRC-013", "SRC-015", "SRC-019", "SRC-021" ], "layers": [ { "id": "dr-rsn-layer-stated-reasons", "name": "Stated reasons and rejection reasons", "description": "The rationale statement issued with a decision, the discrete reason records that compose it, and the contrastive reasons recorded for alternatives that were not selected.", "source_refs": [ "SRC-021", "SRC-022", "SRC-015" ], "findings": [ { "id": "dr-rsn-finding-rationale-statement", "name": "Rationale statement and constituent reasons", "description": "The holistic statement of reasons issued with a decision together with the discrete, independently citable reason records that compose it. Each reason carries its type, its polarity relative to the recorded outcome, an optional asserted weight on a named scale, and the party asserting it. The statement explains the decision; it does not restate the decision problem, option set or criteria, which the `core` split owns.", "source_refs": [ "SRC-021", "SRC-015", "SRC-022" ], "questions": [ { "id": "dr-rsn-q-rationale-vs-reason", "text": "What distinguishes the rationale statement issued with a decision from the individual reason records that compose it?", "kind": "definition", "answer_data": [ "rationale summary text with language tag", "ordered list of constituent reason record identifiers", "scope note stating what the summary does and does not cover" ] }, { "id": "dr-rsn-q-reason-identity", "text": "How is each reason identified so that it can be cited independently of the decision narrative?", "kind": "identity", "answer_data": [ "reason record identifier", "identifier scheme and issuing authority", "master-system identifier of the decision the reason explains" ] }, { "id": "dr-rsn-q-reason-type-polarity", "text": "What type does each reason carry, and does it argue for or against the recorded outcome?", "kind": "classification", "answer_data": [ "reason type code", "reference to the governing reason taxonomy", "polarity value: supporting, opposing or qualifying" ] }, { "id": "dr-rsn-q-reason-weight-scale", "text": "What relative weight or priority is asserted for a reason, and on which named scale is it expressed?", "kind": "measurement", "answer_data": [ "asserted weight value", "scale identifier and its published definition", "note on whether the weight was elicited, derived or unstated" ] }, { "id": "dr-rsn-q-reason-assertor", "text": "Which party asserted the reason, and on what stated basis of authority?", "kind": "authority", "answer_data": [ "reference to the asserting party", "reference to the authority basis or mandate cited", "time the reason was asserted" ] } ], "data_elements": [ { "id": "dr-rsn-de-rationale-summary", "name": "Rationale summary", "description": "Narrative statement of the reasons on which the decision is based, as issued with the decision.", "value_kind": "text", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-021" ] }, { "id": "dr-rsn-de-reason-statement", "name": "Reason statement", "description": "A single discrete reason offered for or against the recorded outcome, individually identified and citable.", "value_kind": "text", "cardinality": "1..n", "required": true, "source_refs": [ "SRC-021", "SRC-015" ] }, { "id": "dr-rsn-de-reason-polarity", "name": "Reason polarity", "description": "Whether the reason supports, opposes or merely qualifies the recorded outcome.", "value_kind": "code", "cardinality": "1", "required": true, "source_refs": [ "SRC-015" ] }, { "id": "dr-rsn-de-reason-weight", "name": "Asserted reason weight", "description": "Relative weight or priority asserted for the reason, meaningful only with its scale reference.", "value_kind": "quantity", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-022" ] }, { "id": "dr-rsn-de-reason-assertor-ref", "name": "Asserting party reference", "description": "Reference to the party that asserted the reason; the party's role assignment is governed elsewhere.", "value_kind": "reference", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-014", "SRC-021" ] } ], "artifacts": [ { "id": "dr-rsn-art-statement-of-reasons", "name": "Statement of reasons", "description": "The issued statement of reasons for a decision, generated from the canonical reason record set and carrying the rationale revision it renders.", "media_or_form": [ "structured reason record set", "narrative statement text", "rendered section of a decision notice" ], "serial": false, "identity_strategy": "Derived from the authoritative master-system identifier of the decision plus a stable rationale-part key; the issue date is metadata and is never used as the identifier.", "source_refs": [ "SRC-021", "SRC-015" ] } ], "inline_only_rationale": null }, { "id": "dr-rsn-finding-rejection-reason", "name": "Reasons recorded for rejecting alternatives", "description": "Contrastive reasons attached to alternatives that were considered but not selected: the ground of rejection, any recorded condition whose change would have made the alternative selectable, and whether the rejection is stated as final for this decision or as a deferral. The alternatives themselves, their definitions and their evaluation against criteria are owned by the `core` split and are referenced, not restated.", "source_refs": [ "SRC-022", "SRC-015", "SRC-021" ], "questions": [ { "id": "dr-rsn-q-rejection-target", "text": "Which non-selected alternative does a rejection reason attach to, without restating that alternative's definition?", "kind": "relationship", "answer_data": [ "reference to the alternative in the `core` split", "reason record identifier", "note confirming the alternative definition is held by reference only" ] }, { "id": "dr-rsn-q-rejection-ground", "text": "On what ground was the alternative not selected: infeasible, dominated, non-compliant, outside appetite or deferred?", "kind": "classification", "answer_data": [ "rejection ground code", "reference to the governing ground vocabulary", "free-text elaboration where no code applies" ] }, { "id": "dr-rsn-q-rejection-reversal", "text": "What condition, if it had been different, is recorded as having made the rejected alternative selectable?", "kind": "constraint", "answer_data": [ "recorded reversal condition text", "reference to the input or assumption that condition depends on", "flag for whether the condition was asserted or left unstated" ] }, { "id": "dr-rsn-q-rejection-finality", "text": "Does the record qualify the rejection as final for this decision or as deferred for later reconsideration?", "kind": "state", "answer_data": [ "rejection scope qualifier", "text of any reconsideration trigger stated in the reason", "note that this qualifier annotates the reason, not the decision lifecycle" ] } ], "data_elements": [ { "id": "dr-rsn-de-rejected-alternative-ref", "name": "Rejected alternative reference", "description": "Reference to the alternative, owned by the `core` split, that the rejection reason explains.", "value_kind": "reference", "cardinality": "1..n", "required": true, "source_refs": [ "SRC-022" ] }, { "id": "dr-rsn-de-rejection-ground", "name": "Rejection ground", "description": "Coded ground on which the alternative was not selected.", "value_kind": "code", "cardinality": "1", "required": true, "source_refs": [ "SRC-022", "SRC-015" ] }, { "id": "dr-rsn-de-reversal-condition", "name": "Recorded reversal condition", "description": "Condition explicitly recorded as one whose change would have made the rejected alternative selectable.", "value_kind": "text", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-022" ] }, { "id": "dr-rsn-de-rejection-scope", "name": "Rejection scope qualifier", "description": "Whether the rejection is stated as final for this decision or as a deferral, as an annotation on the reason.", "value_kind": "code", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-015" ] } ], "artifacts": [], "inline_only_rationale": "Rejection reasons are reference-and-qualifier records that hang off alternatives owned by the `core` split. Materialising them as a separate artefact would create a second place where an alternative's definition could be restated and drift from the governed option set, and would invite consumers to read a rejection note as an independent evaluation. They are therefore carried as inline bindings on the referenced alternative and surface only inside the statement of reasons and explanation renditions produced elsewhere in this model." } ] }, { "id": "dr-rsn-layer-argument-inference", "name": "Argument structure and inference", "description": "How premises, conclusions, inference-scheme references, critical questions and asserted conflict relations are represented, and how explicitly asserted causal or counterfactual reasoning is recorded without being upgraded beyond what was stated.", "source_refs": [ "SRC-013", "SRC-015", "SRC-019" ], "findings": [ { "id": "dr-rsn-finding-argument-structure", "name": "Premises, conclusions and inference-scheme references", "description": "The argument connecting reasons to the recorded outcome, represented with the inference itself as a first-class element rather than as an implicit edge. Covers the role each element plays (premise, conclusion, context, exception), the inference scheme referenced for each step, recorded critical questions and their responses, the support status of each premise, and asserted conflict relations. The model records asserted inference only and performs no validity or entailment evaluation.", "source_refs": [ "SRC-013", "SRC-015", "SRC-019" ], "questions": [ { "id": "dr-rsn-q-argument-composition", "text": "How are premises and conclusions distinguished within a recorded argument, and how is the inference between them represented as a first-class element?", "kind": "composition", "answer_data": [ "argument element role code per node", "identifier of the inference element linking premises to conclusion", "reference to any reasoning description attached to the inference" ] }, { "id": "dr-rsn-q-scheme-reference", "text": "Which inference scheme is referenced for each inference step, and where is that scheme catalogue governed?", "kind": "classification", "answer_data": [ "inference scheme identifier or IRI", "reference to the governing scheme catalogue and its version", "flag for schemes stated informally with no catalogue entry" ] }, { "id": "dr-rsn-q-critical-questions", "text": "Which critical questions or exceptions attached to the referenced scheme are recorded, and are they answered?", "kind": "validation", "answer_data": [ "critical question identifier from the scheme", "recorded response text or explicit unanswered marker", "reference to evidence or claim cited in the response" ] }, { "id": "dr-rsn-q-support-versus-conflict", "text": "How are asserted support relations distinguished from asserted conflict relations between argument elements?", "kind": "relationship", "answer_data": [ "relation type code distinguishing support from conflict", "source and target element identifiers", "note on any asserted preference between conflicting elements" ] }, { "id": "dr-rsn-q-premise-support-status", "text": "Is a given premise recorded as assumed, as axiomatic, or as still requiring support?", "kind": "state", "answer_data": [ "premise support status code", "reference to supporting claim or evidence binding where support exists", "note recording that this status annotates the premise, not the decision" ] } ], "data_elements": [ { "id": "dr-rsn-de-argument-element-role", "name": "Argument element role", "description": "Role an element plays in one argument: premise, conclusion, context, exception or reasoning description.", "value_kind": "code", "cardinality": "1", "required": true, "source_refs": [ "SRC-013", "SRC-015", "SRC-019" ] }, { "id": "dr-rsn-de-inference-scheme-ref", "name": "Inference scheme reference", "description": "Reference to the scheme, in a governed catalogue where one exists, that the asserted inference instantiates.", "value_kind": "reference", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-013" ] }, { "id": "dr-rsn-de-critical-question-response", "name": "Critical question response", "description": "Recorded response to a critical question or exception attached to the referenced scheme, or an explicit unanswered marker.", "value_kind": "text", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-013", "SRC-019" ] }, { "id": "dr-rsn-de-premise-support-status", "name": "Premise support status", "description": "Whether a premise is assumed, axiomatic or still to be supported.", "value_kind": "code", "cardinality": "1", "required": true, "source_refs": [ "SRC-015", "SRC-019" ] }, { "id": "dr-rsn-de-conflict-assertion-ref", "name": "Conflict assertion reference", "description": "Reference to an asserted conflict or attack relation between two argument elements, with any asserted preference between them.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-013", "SRC-015" ] } ], "artifacts": [ { "id": "dr-rsn-art-argument-structure", "name": "Argument structure export", "description": "A projection of the canonical argument record set as a node-and-link structure suitable for exchange with argumentation or assurance-case tooling.", "media_or_form": [ "node-and-link argument graph", "assurance-case argument package", "tabular premise-conclusion-inference listing" ], "serial": false, "identity_strategy": "Derived from the decision's master-system identifier plus the argument key; where the export is exchanged externally it additionally carries the governed scheme-catalogue IRIs it references.", "source_refs": [ "SRC-013", "SRC-015" ] } ], "inline_only_rationale": null }, { "id": "dr-rsn-finding-causal-counterfactual", "name": "Asserted causal and counterfactual reasoning", "description": "Records causal and counterfactual reasoning only where the decision record explicitly asserts it: the causal statement, the strength qualifier used, the antecedent varied in a counterfactual and the outcome claimed to follow, the evidential basis cited, and any competing explanation or confounding considered. The model never upgrades an observed association into an asserted cause, and no consulted standard supplies a normative structure for counterfactual rationale, so this finding is carried as an assertion record with declared limits.", "source_refs": [ "SRC-020", "SRC-016", "SRC-013" ], "questions": [ { "id": "dr-rsn-q-causal-vs-association", "text": "Does the record explicitly assert a causal relation, or does it assert only an association observed in the cited material?", "kind": "definition", "answer_data": [ "causal assertion text as recorded", "assertion mode: causal, associational or unstated", "note on any wording the record uses to hedge the assertion" ] }, { "id": "dr-rsn-q-counterfactual-antecedent", "text": "For an asserted counterfactual, which antecedent is varied and what outcome is claimed to follow from varying it?", "kind": "constraint", "answer_data": [ "counterfactual antecedent statement", "claimed consequent statement", "reference to the assumption or input the antecedent varies" ] }, { "id": "dr-rsn-q-causal-basis", "text": "What evidential basis is cited for the causal assertion, and is that basis experimental or observational?", "kind": "evidence", "answer_data": [ "reference to the evidence citation binding used", "study design descriptor carried from the referenced evidence", "explicit marker where no basis is cited" ] }, { "id": "dr-rsn-q-competing-explanation", "text": "What does the record state about competing explanations or confounding that were considered before the causal claim was made?", "kind": "quality", "answer_data": [ "competing explanation note", "reference to any downgrading domain carried from the evidence appraisal", "marker for causal assertions recorded with no competing-explanation discussion" ] } ], "data_elements": [ { "id": "dr-rsn-de-causal-assertion", "name": "Causal assertion", "description": "Verbatim causal statement as asserted in the decision record.", "value_kind": "text", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-020" ] }, { "id": "dr-rsn-de-causal-strength-qualifier", "name": "Causal strength qualifier", "description": "Qualifier the record attaches to the causal assertion, distinguishing causal claims from associational ones.", "value_kind": "code", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-020", "SRC-016" ] }, { "id": "dr-rsn-de-counterfactual-antecedent", "name": "Counterfactual antecedent", "description": "The antecedent condition varied in an explicitly asserted counterfactual, with its claimed consequent.", "value_kind": "text", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-016" ] }, { "id": "dr-rsn-de-competing-explanation-note", "name": "Competing explanation note", "description": "Recorded consideration of alternative explanations, confounding or residual bias relevant to the causal assertion.", "value_kind": "text", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-020" ] } ], "artifacts": [], "inline_only_rationale": "Causal and counterfactual assertions have no artefact of their own: they are qualifiers on reasons and premises already carried elsewhere in this model, and no consulted primary standard defines a normative document form for them. Producing a separate artefact would give an asserted causal claim standing independent of the reason it qualifies and would invite it to be cited as an established finding. The content is therefore kept inline on the reason or premise record, marked as asserted, and this finding is flagged as an evidence gap in coverage rather than presented as canonical structure." } ] } ] }, { "id": "dr-rsn-bundle-grounding", "name": "Grounding, appraisal and epistemic qualification", "description": "How rationale elements bind to governed claims and evidence, how the decision authority's own appraisal of that evidence is recorded separately from the evidence body's certainty grade, and how assumptions, dependencies, uncertainty, trade-offs, sensitivity and contestation are qualified.", "rationale": "Every consulted framework separates the evidence from the appraisal of it and from the reason it is offered for: certainty-of-evidence frameworks grade a body of evidence, citation ontologies type the rhetorical act of citing, and regulatory-analysis guidance requires assumptions and sensitivity to be disclosed separately from the recommendation. This bundle carries the bindings and epistemic qualifiers required to explain a decision without absorbing the evidence or claim models.", "source_refs": [ "SRC-018", "SRC-020", "SRC-022", "SRC-023" ], "layers": [ { "id": "dr-rsn-layer-bindings", "name": "Claim and evidence bindings", "description": "Reference bindings from rationale elements to governed claims in WM-KNW-007 and governed evidence in WM-KNW-008, carrying the role, citation type, locator and stance but never a copy of the referenced content.", "source_refs": [ "SRC-015", "SRC-018", "SRC-023" ], "findings": [ { "id": "dr-rsn-finding-claim-binding", "name": "Claim references used as premises and conclusions", "description": "Bindings that place a governed claim into an argument role, pin the claim version relied on when the reasoning was asserted, and mark any local paraphrase as a restatement. Claim identity, wording, revision and verification status remain in WM-KNW-007; this finding owns only the binding and the role the claim plays in one argument.", "source_refs": [ "SRC-015", "SRC-013", "SRC-014" ], "questions": [ { "id": "dr-rsn-q-claim-reference-identity", "text": "Which governed claim does each premise or conclusion slot reference, and by which identifier?", "kind": "identity", "answer_data": [ "claim identifier in WM-KNW-007", "identifier scheme used for the reference", "resolution status of the reference at binding time" ] }, { "id": "dr-rsn-q-claim-role", "text": "What role does a referenced claim play in this argument: premise, conclusion, context or exception?", "kind": "composition", "answer_data": [ "claim role code within the argument", "identifier of the argument element occupying that role", "note where one claim occupies multiple roles across arguments" ] }, { "id": "dr-rsn-q-claim-restatement", "text": "How is a locally paraphrased restatement of a claim distinguished from the governed claim it references?", "kind": "definition", "answer_data": [ "local restatement text with an explicit restatement marker", "referenced claim identifier the restatement stands for", "flag indicating whether the restatement was reviewed against the source claim" ] }, { "id": "dr-rsn-q-claim-version-pin", "text": "How does this record express which claim version the reasoning relied on, given that claim revision is governed elsewhere?", "kind": "temporal", "answer_data": [ "pinned claim version or revision identifier", "time the binding was asserted", "note describing how a later claim revision is reported rather than silently followed" ] } ], "data_elements": [ { "id": "dr-rsn-de-claim-ref", "name": "Claim reference", "description": "Reference to a governed claim in WM-KNW-007 used as a premise, conclusion, context or exception.", "value_kind": "reference", "cardinality": "1..n", "required": true, "source_refs": [ "SRC-015" ] }, { "id": "dr-rsn-de-claim-role", "name": "Claim role in argument", "description": "Role the referenced claim occupies within this argument.", "value_kind": "code", "cardinality": "1", "required": true, "source_refs": [ "SRC-015", "SRC-013" ] }, { "id": "dr-rsn-de-claim-version-pin", "name": "Claim version pin", "description": "Version or revision of the referenced claim that the reasoning relied on at the time it was asserted.", "value_kind": "identifier", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-014" ] }, { "id": "dr-rsn-de-local-restatement", "name": "Local claim restatement", "description": "Paraphrase of a referenced claim recorded for readability, explicitly marked as a restatement and never authoritative.", "value_kind": "text", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-015" ] } ], "artifacts": [], "inline_only_rationale": "A claim binding is pure reference data: an identifier, a version pin, a role code and an optional marked paraphrase. Giving it an artefact would produce a second store of claim text outside WM-KNW-007 and create exactly the duplication the reference relation exists to prevent, since a materialised binding document would age independently of the governed claim. The binding therefore lives inline on the argument element and is projected only inside the argument structure export owned by the argument-structure finding." }, { "id": "dr-rsn-finding-evidence-citation", "name": "Evidence citation bindings", "description": "Bindings from a reason or premise to a governed evidence item in WM-KNW-008, carrying the citation type drawn from a governed vocabulary, a locator pinpointing the relevant part of the cited resource, the stance of the citation, and whether the citation was made by the decision author or inherited from an intermediate source. Evidence content, acquisition, custody and grading remain in WM-KNW-008.", "source_refs": [ "SRC-018", "SRC-023", "SRC-015" ], "questions": [ { "id": "dr-rsn-q-evidence-reference", "text": "Which governed evidence item does a citation bind to, and how is an unresolvable reference recorded?", "kind": "identity", "answer_data": [ "evidence identifier in WM-KNW-008", "resolution outcome at binding time", "gap record where the reference does not resolve" ] }, { "id": "dr-rsn-q-citation-type", "text": "What is the type of the citation act: cited as evidence, cited as authority, cited as quoted, disagreed with or refuted?", "kind": "classification", "answer_data": [ "citation type code drawn from a governed citation vocabulary", "vocabulary identifier and version", "free-text note where no vocabulary term fits" ] }, { "id": "dr-rsn-q-citation-locator", "text": "How is the specific part of the cited resource pinpointed rather than the resource being cited as a whole?", "kind": "evidence", "answer_data": [ "locator object with selector kind and parameters", "exact quotation with prefix and suffix context where text is cited", "position range or media fragment where applicable" ] }, { "id": "dr-rsn-q-citation-stance", "text": "Does the citation support the reason it is attached to, counter it, or supply context only?", "kind": "relationship", "answer_data": [ "citation stance code", "identifier of the reason or premise the citation attaches to", "note where the same evidence is cited on both sides" ] }, { "id": "dr-rsn-q-citation-provenance", "text": "Was the citation asserted by the decision author, or inherited from an intermediate document that cited the source?", "kind": "provenance", "answer_data": [ "citation origin mode: direct or inherited", "reference to the intermediate document where inherited", "agent that recorded the citation and the time it was recorded" ] } ], "data_elements": [ { "id": "dr-rsn-de-evidence-ref", "name": "Evidence reference", "description": "Reference to a governed evidence item in WM-KNW-008 cited by a reason or premise.", "value_kind": "reference", "cardinality": "1..n", "required": true, "source_refs": [ "SRC-018", "SRC-015" ] }, { "id": "dr-rsn-de-citation-type", "name": "Citation type", "description": "Coded type of the citation act, drawn from a governed citation-typing vocabulary.", "value_kind": "code", "cardinality": "1", "required": true, "source_refs": [ "SRC-018" ] }, { "id": "dr-rsn-de-citation-locator", "name": "Citation locator", "description": "Selector pinpointing the relevant part of the cited resource, such as a quotation with context, a position range or a media fragment.", "value_kind": "object", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-023" ] }, { "id": "dr-rsn-de-citation-stance", "name": "Citation stance", "description": "Whether the citation supports, counters or contextualises the reason it is attached to.", "value_kind": "code", "cardinality": "1", "required": true, "source_refs": [ "SRC-018", "SRC-015" ] } ], "artifacts": [], "inline_only_rationale": "A citation binding must remain a reference, a type code, a locator and a stance. Materialising it as an artefact would tempt implementers to embed the cited excerpt and thereby create an uncontrolled copy of evidence whose identity, custody and grading belong to WM-KNW-008, breaking the reference contract and producing a second evidence store that can drift or outlive its source. Locators are held inline precisely so that resolution always goes back to the governed evidence item." } ] }, { "id": "dr-rsn-layer-appraisal-uncertainty", "name": "Appraisal, assumptions and confidence", "description": "The decision-local appraisal of cited evidence, the assumptions and dependencies the reasoning rests on, and the confidence and uncertainty qualifiers attached to reasons, inferences and the rationale as a whole.", "source_refs": [ "SRC-020", "SRC-022", "SRC-016", "SRC-019" ], "findings": [ { "id": "dr-rsn-finding-evidential-appraisal", "name": "Decision-local appraisal of cited evidence", "description": "The decision authority's own judgement of how far cited evidence bears on this decision: the weight given to it here, the sufficiency judgement recorded for the reason it supports, and the stated preference where bodies of evidence conflict. The evidence item's intrinsic certainty grade and the grading methodology belong to WM-KNW-008 and are carried by reference; this finding never re-grades evidence and never asserts that the appraisal is correct.", "source_refs": [ "SRC-020", "SRC-015", "SRC-019" ], "questions": [ { "id": "dr-rsn-q-local-weight", "text": "What weight did the decision authority give this evidence for this decision, and how is that kept separate from the evidence item's own certainty grade?", "kind": "measurement", "answer_data": [ "decision-local weight value with its scale", "referenced certainty grade carried from WM-KNW-008", "explicit statement that the two values have different owners" ] }, { "id": "dr-rsn-q-grade-carriage", "text": "Which certainty or quality grade is carried by reference from the evidence model, and which appraisal is asserted locally?", "kind": "quality", "answer_data": [ "referenced certainty grade value and its grading framework", "locally asserted appraisal statement", "timestamp at which the referenced grade was read" ] }, { "id": "dr-rsn-q-sufficiency-basis", "text": "On what stated basis was the cited evidence judged sufficient or insufficient to support the reason it is cited for?", "kind": "evidence", "answer_data": [ "sufficiency judgement code", "basis statement naming the domains or considerations relied on", "reference to the reason whose support is being judged" ] }, { "id": "dr-rsn-q-conflicting-evidence", "text": "Where cited evidence conflicts, how does the record state which body of evidence was preferred and why?", "kind": "decision", "answer_data": [ "references to the conflicting evidence bindings", "preference statement with its stated ground", "note where the conflict was recorded but left unresolved" ] } ], "data_elements": [ { "id": "dr-rsn-de-appraisal-statement", "name": "Appraisal statement", "description": "The decision authority's stated judgement of how the cited evidence bears on this decision.", "value_kind": "text", "cardinality": "1", "required": true, "source_refs": [ "SRC-020", "SRC-019" ] }, { "id": "dr-rsn-de-local-evidential-weight", "name": "Decision-local evidential weight", "description": "Weight assigned to the cited evidence for this decision, on a named local scale.", "value_kind": "quantity", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-020" ] }, { "id": "dr-rsn-de-referenced-certainty-grade", "name": "Referenced certainty grade", "description": "Certainty or quality grade read from WM-KNW-008 and carried unchanged, with the framework and read time that qualify it.", "value_kind": "code", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-020" ] }, { "id": "dr-rsn-de-sufficiency-judgement", "name": "Sufficiency judgement", "description": "Whether the cited evidence was judged sufficient, insufficient or indeterminate for the reason it supports.", "value_kind": "code", "cardinality": "1", "required": true, "source_refs": [ "SRC-020", "SRC-015" ] } ], "artifacts": [], "inline_only_rationale": "The appraisal is a decision-local qualifier on an existing citation binding, not a document. Emitting it as an artefact would produce something that reads like an evidence assessment report and would compete with the grading that WM-KNW-008 owns, blurring exactly the distinction this finding exists to preserve between the evidence, its intrinsic grade and one decision's judgement of its bearing. Keeping it inline forces every reader to encounter the appraisal attached to the specific reason it qualifies." }, { "id": "dr-rsn-finding-assumption-dependency", "name": "Assumptions, dependencies and defeat conditions", "description": "Statements the reasoning rests on that are recorded as assumed rather than supported, the external conditions, inputs and third-party commitments the reasoning depends on, the conditions under which an assumption would be considered defeated, the period or context for which it is asserted to hold, and the party recorded as responsible for its continued validity. This model records assumptions; it does not monitor, test in production or enforce them.", "source_refs": [ "SRC-022", "SRC-019", "SRC-016" ], "questions": [ { "id": "dr-rsn-q-assumption-versus-premise", "text": "Which statements are recorded as assumptions rather than as supported premises, and how is that marking made explicit?", "kind": "definition", "answer_data": [ "assumption statement text", "explicit assumption marker distinguishing it from a supported premise", "reference to the argument element the assumption qualifies" ] }, { "id": "dr-rsn-q-dependency-target", "text": "On what external condition, input, dataset or third-party commitment does the recorded reasoning depend?", "kind": "relationship", "answer_data": [ "reference to the depended-on condition, input or commitment", "dependency type code", "note on whether the dependency was verified at decision time" ] }, { "id": "dr-rsn-q-defeat-condition", "text": "Under what stated condition would an assumption be considered defeated?", "kind": "constraint", "answer_data": [ "defeat condition text", "reference to the observable or input whose change triggers it", "consequence the record states would follow for the affected argument" ] }, { "id": "dr-rsn-q-assumption-validity-window", "text": "For what stated period or context is the assumption asserted to hold?", "kind": "temporal", "answer_data": [ "validity window expressed with explicit start and end where bounded", "context qualifier limiting where the assumption applies", "marker for assumptions asserted without any stated bound" ] }, { "id": "dr-rsn-q-assumption-responsibility", "text": "Who is recorded as responsible for an assumption's continued validity, given that this model performs no monitoring?", "kind": "authority", "answer_data": [ "reference to the responsible party", "reference to the monitoring or review arrangement held outside this model", "explicit statement that no monitoring function is exposed here" ] } ], "data_elements": [ { "id": "dr-rsn-de-assumption-statement", "name": "Assumption statement", "description": "A statement the reasoning rests on that is recorded as assumed rather than supported.", "value_kind": "text", "cardinality": "1..n", "required": true, "source_refs": [ "SRC-019", "SRC-022" ] }, { "id": "dr-rsn-de-dependency-ref", "name": "Dependency reference", "description": "Reference to an external condition, input or third-party commitment the recorded reasoning depends on.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-022" ] }, { "id": "dr-rsn-de-defeat-condition", "name": "Defeat condition", "description": "Condition whose occurrence the record states would defeat the assumption and the argument resting on it.", "value_kind": "text", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-019" ] }, { "id": "dr-rsn-de-assumption-validity-window", "name": "Assumption validity window", "description": "Period or context for which the assumption is asserted to hold, with explicit bounds where stated.", "value_kind": "duration", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-022", "SRC-016" ] }, { "id": "dr-rsn-de-assumption-responsible-ref", "name": "Assumption responsibility reference", "description": "Reference to the party recorded as responsible for the assumption's continued validity; monitoring is executed elsewhere.", "value_kind": "reference", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-022" ] } ], "artifacts": [ { "id": "dr-rsn-art-assumption-register", "name": "Assumption and dependency register", "description": "The consolidated register of assumptions, dependencies, defeat conditions and validity windows for one decision, generated from the canonical assumption records so that a third party can see what the reasoning rests on.", "media_or_form": [ "structured register rows", "tabular listing", "rendered annex to the decision record" ], "serial": false, "identity_strategy": "Register identity derives from the decision's master-system identifier plus the register key; individual rows use a governed identifier where one exists and otherwise a ULID assigned by the adopting Dimension.", "source_refs": [ "SRC-022", "SRC-019" ] } ], "inline_only_rationale": null }, { "id": "dr-rsn-finding-confidence-uncertainty", "name": "Uncertainty and confidence qualifiers", "description": "Confidence qualifiers attached to reasons, inferences and the rationale as a whole: the scale used, whether the qualifier is about the evidence, the inference or the outcome holding in practice, the stated limits of knowledge beyond which the reasoning should not be relied on, and how a verbal qualifier maps to any numeric range. The model records the qualifier as asserted and does not compute or validate calibration.", "source_refs": [ "SRC-020", "SRC-016", "SRC-022" ], "questions": [ { "id": "dr-rsn-q-confidence-scale", "text": "What confidence qualifier is attached to the rationale as a whole, and on which named and published scale is it expressed?", "kind": "measurement", "answer_data": [ "confidence qualifier value", "scale identifier with its published level definitions", "the element the qualifier is attached to" ] }, { "id": "dr-rsn-q-confidence-subject", "text": "Is the qualifier a statement about the evidence, about the inference, or about the outcome holding in practice?", "kind": "definition", "answer_data": [ "confidence subject code", "reference to the qualified element", "note where a single qualifier was asserted over several subjects at once" ] }, { "id": "dr-rsn-q-knowledge-limits", "text": "What are the stated limits of knowledge beyond which the recorded reasoning should not be relied on?", "kind": "constraint", "answer_data": [ "knowledge limit statement", "conditions or populations explicitly excluded from the reasoning", "reference to the assumption records that bound the reasoning" ] }, { "id": "dr-rsn-q-qualifier-mapping", "text": "How is a calibrated verbal qualifier mapped to a numeric range, and is that mapping published with the record?", "kind": "interoperability", "answer_data": [ "verbal-to-numeric mapping table reference", "publication location of the scale definition", "flag where a verbal qualifier is used with no published mapping" ] } ], "data_elements": [ { "id": "dr-rsn-de-confidence-qualifier", "name": "Confidence qualifier", "description": "Asserted confidence value attached to a reason, inference or the rationale as a whole.", "value_kind": "code", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-020" ] }, { "id": "dr-rsn-de-confidence-scale-ref", "name": "Confidence scale reference", "description": "Reference to the named scale whose published level definitions give the qualifier meaning.", "value_kind": "reference", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-020" ] }, { "id": "dr-rsn-de-confidence-subject", "name": "Confidence subject", "description": "Whether the qualifier speaks to the evidence, the inference or the outcome holding in practice.", "value_kind": "code", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-020", "SRC-016" ] }, { "id": "dr-rsn-de-knowledge-limit-statement", "name": "Knowledge limit statement", "description": "Stated conditions beyond which the recorded reasoning is declared not to apply.", "value_kind": "text", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-016" ] } ], "artifacts": [], "inline_only_rationale": "Confidence is a qualifier on something else — a reason, an inference, an appraisal or the whole rationale — and has no standing apart from what it qualifies. Extracting qualifiers into their own artefact would detach a confidence level from the statement it modifies, which is the specific failure mode calibrated-language guidance exists to prevent, and would let a summary confidence figure circulate without its scale definition or knowledge limits. Qualifiers are therefore inline and always travel with their subject and scale reference." } ] }, { "id": "dr-rsn-layer-robustness-contestation", "name": "Robustness and contestation", "description": "Recorded trade-offs and sensitivity of the reasoning to its inputs, and the counterarguments, counter-evidence, objections, dispositions and dissenting or minority positions recorded against it.", "source_refs": [ "SRC-022", "SRC-019", "SRC-015" ], "findings": [ { "id": "dr-rsn-finding-tradeoff-sensitivity", "name": "Trade-offs and sensitivity of the reasoning", "description": "What was traded against what, which disbenefits are recorded as accepted, which inputs or assumptions would change the recorded outcome and within what range, any switching value at which the preferred option would change, and whether the varied inputs are disclosed well enough for a qualified third party to reproduce the finding. This model records the results and the disclosure; it does not run the analysis.", "source_refs": [ "SRC-022", "SRC-020" ], "questions": [ { "id": "dr-rsn-q-tradeoff-composition", "text": "Which benefits and disbenefits were traded against each other, and which disbenefits are recorded as knowingly accepted?", "kind": "composition", "answer_data": [ "trade-off statement naming both sides", "accepted disbenefit entries", "references to the criteria in the `core` split that each side maps to" ] }, { "id": "dr-rsn-q-sensitivity-variables", "text": "Which inputs or assumptions, if changed, are recorded as changing the outcome, and over what stated range were they varied?", "kind": "measurement", "answer_data": [ "sensitivity variable references", "range or set of values over which each was varied", "recorded effect on the outcome for each variation" ] }, { "id": "dr-rsn-q-switching-value", "text": "What switching value or threshold is recorded at which the preferred option would change?", "kind": "constraint", "answer_data": [ "switching value with units and the variable it applies to", "the option that would become preferred beyond it", "note where no switching value could be identified" ] }, { "id": "dr-rsn-q-sensitivity-reproducibility", "text": "Are the varied inputs and method disclosed well enough for a qualified third party to see how the result was reached?", "kind": "validation", "answer_data": [ "disclosure statement covering inputs, method and data sources", "reference to the analysis record or dataset held outside this model", "gap marker where disclosure is incomplete" ] }, { "id": "dr-rsn-q-no-sensitivity-analysis", "text": "Does the record state that no sensitivity analysis was performed, and what reason is given?", "kind": "exception", "answer_data": [ "explicit no-analysis marker", "stated reason for omission", "reference to any waiver or proportionality argument relied on" ] } ], "data_elements": [ { "id": "dr-rsn-de-tradeoff-statement", "name": "Trade-off statement", "description": "Statement of what was traded against what in reaching the recorded outcome.", "value_kind": "text", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-022" ] }, { "id": "dr-rsn-de-accepted-disbenefit", "name": "Accepted disbenefit", "description": "A disbenefit the record states was knowingly accepted as part of the trade-off.", "value_kind": "text", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-022" ] }, { "id": "dr-rsn-de-sensitivity-variable-ref", "name": "Sensitivity variable reference", "description": "Reference to an input or assumption whose variation was tested, with the range over which it was varied.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-022" ] }, { "id": "dr-rsn-de-switching-value", "name": "Switching value", "description": "Value of a variable at which the recorded preference between options would change.", "value_kind": "quantity", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-022" ] }, { "id": "dr-rsn-de-reproducibility-note", "name": "Reproducibility disclosure note", "description": "Statement of inputs, method and sources sufficient for a qualified third party to follow how the result was reached.", "value_kind": "text", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-022" ] } ], "artifacts": [ { "id": "dr-rsn-art-sensitivity-record", "name": "Sensitivity and trade-off record", "description": "The presented record of trade-offs, varied inputs, ranges, outcomes and any switching value for one analysis run supporting the decision.", "media_or_form": [ "tabular sensitivity table", "structured variation set", "narrative robustness note" ], "serial": true, "identity_strategy": "Identified by the decision's master-system identifier plus the artefact-kind key and a zero-padded run sequence; the analysis it reports is referenced by its own identifier and is not owned here.", "source_refs": [ "SRC-022" ] } ], "inline_only_rationale": null }, { "id": "dr-rsn-finding-counterargument-dissent", "name": "Counterarguments, objections and dissent", "description": "Counter-arguments and counter-evidence recorded against a reason or inference, who raised them, how each was dispositioned in the rationale, which argument elements are marked defeated by a recorded challenge, and any minority or dissenting position preserved as an attributed statement rather than folded into the majority rationale. Adjudication authority, approval workflow and role assignment stay with the `governance` split.", "source_refs": [ "SRC-019", "SRC-015", "SRC-013" ], "questions": [ { "id": "dr-rsn-q-counter-target", "text": "What counter-argument or counter-evidence is recorded, and which reason or inference does it attack?", "kind": "relationship", "answer_data": [ "objection statement text", "reference to the attacked reason, premise or inference", "reference to any counter-evidence citation binding" ] }, { "id": "dr-rsn-q-objection-origin", "text": "Who raised the objection and in what stated capacity, without this model assigning their role?", "kind": "provenance", "answer_data": [ "reference to the objecting party", "capacity as stated in the record", "time the objection was raised" ] }, { "id": "dr-rsn-q-objection-disposition", "text": "How was the objection dispositioned in the rationale: accepted, rebutted, undercut or left open?", "kind": "state", "answer_data": [ "disposition code", "disposition statement with its stated ground", "time and asserting party of the disposition" ] }, { "id": "dr-rsn-q-minority-position", "text": "Is a minority or dissenting position preserved as an attributed position rather than merged into the majority rationale?", "kind": "definition", "answer_data": [ "dissenting position statement", "attribution reference for the dissenting party", "explicit marker separating it from the majority rationale" ] }, { "id": "dr-rsn-q-defeated-marking", "text": "Which argument elements are marked as defeated by a recorded challenge, and what does that marking not change?", "kind": "constraint", "answer_data": [ "defeated marking on the affected argument element", "reference to the challenge that produced it", "explicit note that the marking annotates the argument and does not alter the recorded outcome or decision state" ] } ], "data_elements": [ { "id": "dr-rsn-de-objection-statement", "name": "Objection statement", "description": "Counter-argument, objection or counter-evidence recorded against a reason or inference, captured as stated.", "value_kind": "text", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-015", "SRC-019" ] }, { "id": "dr-rsn-de-objection-target-ref", "name": "Objection target reference", "description": "Reference to the reason, premise or inference the objection attacks.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-013", "SRC-015" ] }, { "id": "dr-rsn-de-objection-disposition", "name": "Objection disposition", "description": "How the objection was dealt with in the rationale: accepted, rebutted, undercut or left open.", "value_kind": "code", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-019", "SRC-015" ] }, { "id": "dr-rsn-de-dissent-attribution-ref", "name": "Dissent attribution reference", "description": "Reference identifying the party whose minority or dissenting position is recorded.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-014", "SRC-019" ] }, { "id": "dr-rsn-de-defeated-marking", "name": "Defeated marking", "description": "Annotation marking an argument element as defeated by a recorded challenge; an argument-level marking only.", "value_kind": "boolean", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-019" ] } ], "artifacts": [ { "id": "dr-rsn-art-dissent-record", "name": "Dissenting or minority opinion", "description": "An attributed statement of a minority or dissenting position preserved alongside the majority rationale, together with the objections it rests on.", "media_or_form": [ "attributed narrative statement", "structured objection record", "rendered annex to the decision notice" ], "serial": true, "identity_strategy": "Identified by the decision's master-system identifier plus the artefact-kind key and a zero-padded sequence per dissenting party; attribution references the party record held outside this model.", "source_refs": [ "SRC-019", "SRC-015" ] } ], "inline_only_rationale": null } ] } ] }, { "id": "dr-rsn-bundle-explanation", "name": "Explanation and reasoning provenance", "description": "Explainable decision traces and bounded, audience-scoped explanation renditions produced from the recorded rationale, together with the provenance of the reasoning itself and the separation of authored rationale from machine-generated explanation.", "rationale": "Explainability guidance requires an explanation to be meaningful to its recipient and accurate to the process that generated the output, and EU law gives affected persons a right to clear and meaningful explanations of an AI system's role in a decision. A provenance vocabulary supplies the attribution terms, including software agents, needed to say who or what produced the reasoning. This bundle separates the rendition given to an audience from the reasoning it renders and from the provenance of that reasoning.", "source_refs": [ "SRC-016", "SRC-017", "SRC-014" ], "layers": [ { "id": "dr-rsn-layer-explanation-provenance", "name": "Decision traces, renditions and reasoning provenance", "description": "The ordered trace from inputs to recorded outcome, the bounded renditions produced for declared audiences with their faithfulness and omission statements, and the provenance of the reasoning including authorship mode and generating agent.", "source_refs": [ "SRC-016", "SRC-017", "SRC-014" ], "findings": [ { "id": "dr-rsn-finding-decision-trace", "name": "Explainable decision trace and bounded explanation renditions", "description": "The ordered sequence of reasoning steps recorded as the trace from inputs to the recorded outcome, and the audience-scoped explanations produced from it. Each rendition declares its audience, what was omitted or replaced by a reference, the rationale revision it renders, and an assertion that it reflects the recorded reasoning rather than a post-hoc narrative. A rendition is an explanation, not an audit record: audit trails and access logging are owned by the `governance` split.", "source_refs": [ "SRC-016", "SRC-017", "SRC-014" ], "questions": [ { "id": "dr-rsn-q-trace-steps", "text": "What ordered sequence of reasoning steps is recorded as the trace from inputs to the recorded outcome?", "kind": "process", "answer_data": [ "ordered trace step entries with explicit sequence keys", "reference from each step to the reason, inference or binding it corresponds to", "inputs entering the trace, carried as references" ] }, { "id": "dr-rsn-q-rendition-audience", "text": "For which audience is a rendition produced, and what does the record state was omitted or replaced by a reference for that audience?", "kind": "requirement", "answer_data": [ "audience code and the entitlement or duty the rendition answers", "omission set listing what was withheld and on what stated ground", "reference to the omission policy held by the `governance` split" ] }, { "id": "dr-rsn-q-rendition-faithfulness", "text": "How is a rendition asserted to reflect the recorded reasoning rather than a plausible reconstruction of it?", "kind": "quality", "answer_data": [ "faithfulness assertion with the method used to derive the rendition", "identifier and revision of the rationale record set rendered", "marker where the rendition is a summary rather than a complete restatement" ] }, { "id": "dr-rsn-q-trace-bounds", "text": "What are the declared bounds of the trace: which steps are recorded in full, which are summarised, and which are not captured at all?", "kind": "constraint", "answer_data": [ "trace completeness declaration per step class", "list of step classes explicitly not captured", "reference to the system boundary beyond which steps are not observable" ] }, { "id": "dr-rsn-q-rendition-linkage", "text": "How is each issued rendition identified and linked back to the rationale revision it renders?", "kind": "identity", "answer_data": [ "rendition identifier with its serial sequence", "rationale record set identifier and revision", "content digest of the rendition as issued" ] } ], "data_elements": [ { "id": "dr-rsn-de-trace-step", "name": "Trace step", "description": "One recorded reasoning step in the ordered trace, referencing the reason, inference or binding it corresponds to.", "value_kind": "object", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-016" ] }, { "id": "dr-rsn-de-explanation-audience", "name": "Explanation audience", "description": "Declared audience a rendition is produced for, and the entitlement or duty it answers.", "value_kind": "code", "cardinality": "1", "required": true, "source_refs": [ "SRC-017", "SRC-016" ] }, { "id": "dr-rsn-de-faithfulness-assertion", "name": "Faithfulness assertion", "description": "Assertion that the rendition reflects the recorded reasoning, naming the derivation method and the rationale revision rendered.", "value_kind": "text", "cardinality": "1", "required": true, "source_refs": [ "SRC-016" ] }, { "id": "dr-rsn-de-explanation-omission-note", "name": "Explanation omission note", "description": "Record of what was omitted or replaced by a reference in a rendition and on what stated ground.", "value_kind": "text", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-017" ] } ], "artifacts": [ { "id": "dr-rsn-art-explanation-rendition", "name": "Explanation rendition", "description": "An audience-scoped explanation issued from the recorded rationale, carrying its audience, omission set, faithfulness assertion and the rationale revision it renders.", "media_or_form": [ "structured explanation record", "narrative explanation text", "rendered notice supplied to a recipient" ], "serial": true, "identity_strategy": "Identified by the decision's master-system identifier plus the artefact-kind key and a zero-padded issuance sequence; each issuance additionally carries a content digest and the rendered rationale revision.", "source_refs": [ "SRC-017", "SRC-016" ] }, { "id": "dr-rsn-art-reasoning-trace", "name": "Reasoning trace record", "description": "The ordered trace of reasoning steps from inputs to recorded outcome, with its completeness declaration; a description of reasoning, not an audit log.", "media_or_form": [ "ordered structured step sequence", "tabular step listing", "graph projection over argument elements" ], "serial": false, "identity_strategy": "Derived from the decision's master-system identifier plus the trace key; steps use explicit sequence numbers and never derive identity from their timestamps.", "source_refs": [ "SRC-016", "SRC-014" ] } ], "inline_only_rationale": null }, { "id": "dr-rsn-finding-reasoning-provenance", "name": "Reasoning provenance and authored versus generated rationale", "description": "Who or what produced the recorded reasoning: whether it was authored by a person, generated by a software agent, or generated then edited; which agent, model version and configuration produced any machine-generated text; what prior rationale, template or precedent it was derived from; and the separation of the time the reasoning was asserted from the time the record was observed or ingested. Machine-generated explanation must be labelled so that it is never read as an authored reason.", "source_refs": [ "SRC-014", "SRC-016", "SRC-017" ], "questions": [ { "id": "dr-rsn-q-authorship-mode", "text": "Was the rationale authored by a person, generated by a software agent, or generated and then edited by a person?", "kind": "provenance", "answer_data": [ "authorship mode code", "reference to the responsible human agent where one exists", "reference to the generating activity where machine generation occurred" ] }, { "id": "dr-rsn-q-generating-agent", "text": "Which software agent, model version and configuration produced a machine-generated explanation?", "kind": "identity", "answer_data": [ "software agent identifier", "model or tool version string and configuration reference", "association between the agent and the generating activity" ] }, { "id": "dr-rsn-q-assertion-versus-record-time", "text": "At what time was the reasoning asserted, and at what time was the record of it observed or ingested?", "kind": "temporal", "answer_data": [ "asserted time with explicit seconds and offset", "recorded or ingested time with explicit seconds and offset", "note where the two are equal because the record is the original assertion" ] }, { "id": "dr-rsn-q-rationale-derivation", "text": "From which prior rationale, template or precedent was this reasoning derived?", "kind": "relationship", "answer_data": [ "reference to the source rationale, template or precedent", "derivation type descriptor", "note on what was changed relative to the source" ] }, { "id": "dr-rsn-q-generated-labelling", "text": "How is machine-generated explanation labelled so that a reader cannot mistake it for a reason asserted by the decision authority?", "kind": "requirement", "answer_data": [ "generated-content label applied at every projection", "reference to the human agent who reviewed or adopted the generated text, where any did", "rule statement covering how the label survives export and rendering" ] } ], "data_elements": [ { "id": "dr-rsn-de-authorship-mode", "name": "Rationale authorship mode", "description": "Whether the rationale was authored, machine-generated, or generated and then edited by a person.", "value_kind": "code", "cardinality": "1", "required": true, "source_refs": [ "SRC-014", "SRC-016" ] }, { "id": "dr-rsn-de-generating-agent-ref", "name": "Generating agent reference", "description": "Reference to the software agent, with model version and configuration, that produced machine-generated explanation.", "value_kind": "reference", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-014", "SRC-017" ] }, { "id": "dr-rsn-de-asserted-at-time", "name": "Reasoning asserted time", "description": "Time the reasoning was asserted by its author or generating agent, as an event time.", "value_kind": "timestamp", "cardinality": "1", "required": true, "source_refs": [ "SRC-014" ] }, { "id": "dr-rsn-de-recorded-at-time", "name": "Record observation time", "description": "Time the rationale record was observed or ingested by the rationale store, recorded separately from the asserted time.", "value_kind": "timestamp", "cardinality": "1", "required": true, "source_refs": [ "SRC-014" ] }, { "id": "dr-rsn-de-derived-from-ref", "name": "Derived-from reference", "description": "Reference to a prior rationale, template or precedent from which this reasoning was derived.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-014" ] } ], "artifacts": [], "inline_only_rationale": "Reasoning provenance qualifies every other record in this model rather than standing on its own: authorship mode, generating agent and the two timestamps must travel with the reason, appraisal, objection or rendition they describe, or the labelling requirement for machine-generated content fails exactly where it matters. A separate provenance artefact would also duplicate the agent and activity records that the provenance vocabulary places outside this model, and would let a rendition circulate stripped of its generated-content label. Provenance is therefore carried inline on each record and projected into artefacts as attribution metadata." } ] } ] }, { "id": "dr-gov-bun-authority", "name": "Decision authority, mandate and participation", "description": "Establishes who was competent to take the decision, under which mandate and delegation, who is accountable for the resulting record, who participated in which declared role, and what integrity conditions applied to that participation.", "rationale": "A decision is only authoritative if the deciding body acted within a mandate and the participation conditions prescribed for it. ISO 15489-1 makes assigned responsibilities and mandate-driven records requirements a precondition of authoritative evidence; PROV qualifies responsibility, association and delegation; the Council's Rules of Procedure show mandate, quorum and proxy representation as explicit, recorded facts.", "source_refs": [ "SRC-024", "SRC-014", "SRC-032" ], "layers": [ { "id": "dr-gov-lay-authority-mandate", "name": "Mandate, deciding authority and accountable roles", "description": "The competence chain: the authority that decided, the instrument that empowered it, the delegations exercised beneath it, and the standing roles accountable for the decision record afterwards.", "source_refs": [ "SRC-024", "SRC-014", "SRC-031" ], "findings": [ { "id": "dr-gov-fnd-deciding-authority", "name": "Deciding authority and mandate binding", "description": "Identifies the body, office or role that took the decision and binds it to the mandate instrument, its subject and jurisdictional limits, its validity window and its succession path. Records the authority as an attributable agent rather than as free text.", "source_refs": [ "SRC-024", "SRC-014", "SRC-031", "SRC-032" ], "questions": [ { "id": "dr-gov-q-auth-identity", "text": "Which authoritative master-system identifier designates the body or office that took this decision, and under which mandate instrument did it act?", "kind": "authority", "answer_data": [ "Deciding authority reference resolved against the party or governing-body register", "Mandate instrument reference with its issuing authority", "Statement of which system of record issued the authority identifier" ] }, { "id": "dr-gov-q-auth-limits", "text": "What subject-matter, monetary or jurisdictional limits does the mandate place on this authority, and what is the recorded consequence of acting outside them?", "kind": "constraint", "answer_data": [ "Enumerated mandate scope limits with units where quantified", "Ultra vires handling code and the status the decision takes if a limit is breached", "Reference to the reviewing forum competent to determine the breach" ] }, { "id": "dr-gov-q-auth-window", "text": "Over which validity window was the mandate in force at the instant the decision was taken?", "kind": "temporal", "answer_data": [ "Mandate validity start and end as RFC 3339 timestamps with explicit offset", "Decision event time compared against that window", "Flag indicating whether the comparison was asserted or externally verified" ] }, { "id": "dr-gov-q-auth-succession", "text": "Which alternate or successor authority becomes responsible if the deciding body is dissolved, merged or renamed?", "kind": "ownership", "answer_data": [ "Successor authority reference with the transfer instrument reference", "Effective time of the succession", "Retained historical authority reference for the original decision" ] } ], "data_elements": [ { "id": "dr-gov-de-authority-ref", "name": "Deciding authority reference", "description": "Reference to the body, office or role that took the decision, resolved against the authoritative party or governing-body register.", "value_kind": "reference", "cardinality": "1", "required": true, "source_refs": [ "SRC-014", "SRC-031" ] }, { "id": "dr-gov-de-mandate-ref", "name": "Mandate instrument reference", "description": "Reference to the statute, charter, terms of reference, delegation instrument or standing order that confers competence for this decision.", "value_kind": "reference", "cardinality": "1..n", "required": true, "source_refs": [ "SRC-024", "SRC-031" ] }, { "id": "dr-gov-de-mandate-limit", "name": "Mandate scope limit", "description": "A declared limit on the mandate expressed as subject-matter, monetary ceiling, geography or term.", "value_kind": "object", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-032" ] }, { "id": "dr-gov-de-mandate-window", "name": "Mandate validity window", "description": "Start and end instants during which the mandate was in force, each an RFC 3339 timestamp with explicit offset or Z.", "value_kind": "object", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-026", "SRC-028" ] }, { "id": "dr-gov-de-authority-succession-ref", "name": "Authority succession reference", "description": "Reference to the successor or alternate authority and to the instrument effecting the succession.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-031" ] } ], "artifacts": [ { "id": "dr-gov-art-authority-register", "name": "Decision authority and delegation register", "description": "A governed register of deciding authorities, their mandate instruments, scope limits, validity windows and the delegations issued beneath them, against which any individual decision's authority claim is resolved. It is a register of competence, not a record of decisions taken.", "media_or_form": [ "structured register entries", "tabular listing rendition", "signed or sealed authoritative extract" ], "serial": false, "identity_strategy": "Identity is taken from the authoritative master-system identifier of the governing-body or decision-rights register maintained by the adopting Dimension; failing that, a governed IRI assigned by a recognised registry; failing both, a ULID minted by the adopting Dimension and flagged as Dimension-assigned. Register entries are keyed by authority identifier plus mandate instrument identifier, never by adoption date or meeting name.", "source_refs": [ "SRC-024", "SRC-025", "SRC-031" ] } ], "inline_only_rationale": null }, { "id": "dr-gov-fnd-delegation-limits", "name": "Delegation, sub-delegation and proxy representation", "description": "Records the chain from the mandate holder to the person or system that actually exercised the decision, with the ceiling and validity window on each step, whether sub-delegation was permitted, and how a vote or act exercised on behalf of another participant is captured.", "source_refs": [ "SRC-014", "SRC-032" ], "questions": [ { "id": "dr-gov-q-del-chain", "text": "Which delegation or sub-delegation steps link the mandate holder to the agent that actually exercised this decision?", "kind": "relationship", "answer_data": [ "Ordered delegation chain entries, each naming delegator, delegate and instrument", "Qualified delegation reference aligned to the PROV delegation pattern", "Indicator of whether each step was standing or ad hoc" ] }, { "id": "dr-gov-q-del-ceiling", "text": "What ceiling applies to each delegation step, and is further sub-delegation permitted at that step?", "kind": "constraint", "answer_data": [ "Ceiling value and unit per step", "Sub-delegation permitted flag with the authorising clause reference", "Escalation target when the ceiling is exceeded" ] }, { "id": "dr-gov-q-del-revocation", "text": "How is a delegation revoked or withdrawn, and what is the effect on decisions already taken under it?", "kind": "authority", "answer_data": [ "Revocation instrument reference and effective instant", "Retrospectivity code stating whether prior decisions remain valid", "Ratification requirement flag for acts taken after revocation" ] }, { "id": "dr-gov-q-del-proxy", "text": "When an act or vote was exercised on behalf of another participant, how is that representation recorded?", "kind": "process", "answer_data": [ "Proxy entry naming represented party, representing party and the authorising provision", "Limit on the number of parties one participant may represent", "Whether the proxy counted towards the participation threshold" ] } ], "data_elements": [ { "id": "dr-gov-de-delegation-entry", "name": "Delegation chain entry", "description": "One step of delegated competence, naming delegator, delegate, instrument reference and ordinal position in the chain.", "value_kind": "object", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-014" ] }, { "id": "dr-gov-de-delegation-ceiling", "name": "Delegation ceiling", "description": "The quantified limit attached to a delegation step, with value, unit and basis.", "value_kind": "quantity", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-032" ] }, { "id": "dr-gov-de-delegation-window", "name": "Delegation validity window", "description": "Start and end instants of a delegation step as RFC 3339 timestamps with explicit offset or Z.", "value_kind": "object", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-026", "SRC-028" ] }, { "id": "dr-gov-de-delegation-revocation-ref", "name": "Delegation revocation reference", "description": "Reference to the instrument revoking or withdrawing a delegation, with its retrospectivity code.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-014" ] }, { "id": "dr-gov-de-proxy-entry", "name": "Proxy representation entry", "description": "Record that one participant acted on behalf of another, naming both parties and the authorising provision.", "value_kind": "object", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-032" ] } ], "artifacts": [], "inline_only_rationale": "Delegation facts are short reference structures resolved against dr-gov-art-authority-register and mirrored into the lifecycle ledger; they carry no independent document body. The delegation instrument itself is a document owned and published by the issuing authority, so materialising a second artefact here would duplicate an external record and create a competing copy of a mandate this model does not own." }, { "id": "dr-gov-fnd-accountable-roles", "name": "Owner, steward, custodian and reviewer assignment", "description": "Assigns the standing accountabilities over the decision record after it is taken: who owns it, who maintains its governance metadata, who holds physical or logical custody of its projections, and who is obliged to review it. Distinguishes accountability for governance metadata from accountability for the substance of the decision.", "source_refs": [ "SRC-024", "SRC-025", "SRC-027" ], "questions": [ { "id": "dr-gov-q-role-distinction", "text": "Who is the accountable owner of this decision record, and how does that accountability differ from those of the steward, custodian and reviewer?", "kind": "ownership", "answer_data": [ "Party references for owner, steward, custodian and reviewer", "Responsibility statement per role drawn from the adopting Dimension's role catalogue", "Indicator of whether a role is held by a person, an office or an organisational unit" ] }, { "id": "dr-gov-q-role-metadata-accountability", "text": "Which role is answerable for the accuracy of the governance metadata as distinct from the correctness of the decision itself?", "kind": "quality", "answer_data": [ "Named role accountable for governance metadata accuracy", "Named authority accountable for decision substance", "Escalation path when the two disagree" ] }, { "id": "dr-gov-q-role-handover", "text": "How is a role assignment transferred when the holder leaves, and from which instant does the successor's accountability run?", "kind": "lifecycle", "answer_data": [ "Handover event reference with predecessor and successor party references", "Handover effective instant as an RFC 3339 timestamp with explicit offset or Z", "Whether the predecessor retains residual accountability for the prior period" ] }, { "id": "dr-gov-q-role-release-authority", "text": "Which role holds the authority to release this decision to a wider audience or to publish it?", "kind": "authority", "answer_data": [ "Release authority role reference", "Constraint stating whether release requires the deciding authority's concurrence", "Reference to the access schedule entry that the release must satisfy" ] } ], "data_elements": [ { "id": "dr-gov-de-owner-ref", "name": "Governance owner reference", "description": "Party or office accountable for the decision record as a governed asset.", "value_kind": "reference", "cardinality": "1", "required": true, "source_refs": [ "SRC-024" ] }, { "id": "dr-gov-de-steward-ref", "name": "Steward reference", "description": "Party accountable for the accuracy and completeness of the decision's governance metadata.", "value_kind": "reference", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-024" ] }, { "id": "dr-gov-de-custodian-ref", "name": "Custodian reference", "description": "Party holding custody of a stored projection of the decision record.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-025" ] }, { "id": "dr-gov-de-reviewer-ref", "name": "Reviewer reference", "description": "Party or forum obliged to review the decision at scheduled or triggered intervals.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-029" ] }, { "id": "dr-gov-de-role-handover-entry", "name": "Role handover entry", "description": "Transfer of a named role from predecessor to successor with its effective instant and residual-accountability flag.", "value_kind": "object", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-014", "SRC-028" ] } ], "artifacts": [], "inline_only_rationale": "Role assignments are reference tuples pointing at party identifiers held in an external identity model; they have no document body of their own and change independently of the decision content. Materialising them as an artefact would create a stale local copy of party master data and a second, competing source of truth for accountability that the adopting Dimension's role catalogue already owns." } ] }, { "id": "dr-gov-lay-participation-integrity", "name": "Participation, decision rule and integrity of participation", "description": "Who took part in which declared capacity, under which decision rule and participation threshold, and what was declared, determined and acted upon regarding conflicts of interest.", "source_refs": [ "SRC-012", "SRC-030", "SRC-032" ], "findings": [ { "id": "dr-gov-fnd-participation-decision-rule", "name": "Participant roster, declared roles and applicable decision rule", "description": "Captures the participants and their declared roles, the entitlement to vote or object, the decision rule that governed the taking of the decision, the mode of proceeding, and the attestation that any required participation threshold was met.", "source_refs": [ "SRC-012", "SRC-029", "SRC-032" ], "questions": [ { "id": "dr-gov-q-part-roster", "text": "Which participants took part, in which declared roles, and which of them were entitled to vote or formally object?", "kind": "relationship", "answer_data": [ "Participant entries with party reference and declared role code", "Voting entitlement flag per participant", "Alignment of role codes to a controlled role vocabulary such as a top-level role classification" ] }, { "id": "dr-gov-q-part-rule", "text": "Which decision rule applied — unanimity, qualified or simple majority, consensus, silence or written procedure — and where is that rule prescribed?", "kind": "requirement", "answer_data": [ "Decision rule code from a controlled vocabulary", "Reference to the rules of procedure clause prescribing it", "Threshold expression where the rule is numeric" ] }, { "id": "dr-gov-q-part-quorum", "text": "Was the participation threshold or quorum asserted as satisfied, by whom, and on what stated basis?", "kind": "validation", "answer_data": [ "Quorum attestation with attesting party reference and attestation instant", "Counted and required participation figures as asserted", "Explicit statement that this is the presiding authority's assertion, not an evaluation performed by this model" ] }, { "id": "dr-gov-q-part-mode", "text": "How is a decision taken by written or silence procedure distinguished from one taken in a convened sitting?", "kind": "process", "answer_data": [ "Participation mode code covering convened, written, silence and delegated modes", "Objection window for silence procedure with its closing instant", "Reference to the minutes or proceedings record where one exists" ] }, { "id": "dr-gov-q-part-abstention", "text": "How is an abstention or formal objection referenced without reproducing the dissent record owned by the reasoning area?", "kind": "event", "answer_data": [ "Abstention or objection event entry with participant reference and instant", "Pointer to dr-rsn-art-dissent-record by identifier only", "Rule stating that dissent content, arguments and grounds are not stored here" ] } ], "data_elements": [ { "id": "dr-gov-de-participant-entry", "name": "Participant entry", "description": "One participant with party reference, declared role code, voting entitlement flag and attendance or response state.", "value_kind": "object", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-012", "SRC-032" ] }, { "id": "dr-gov-de-decision-rule-code", "name": "Decision rule code", "description": "Controlled code naming the rule under which the decision was taken, with a reference to the prescribing provision.", "value_kind": "code", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-029", "SRC-032" ] }, { "id": "dr-gov-de-quorum-attestation", "name": "Quorum attestation", "description": "Assertion by a named presiding party that the participation threshold was satisfied, with counted and required figures and the attestation instant.", "value_kind": "object", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-032" ] }, { "id": "dr-gov-de-participation-mode-code", "name": "Participation mode code", "description": "Code distinguishing convened sitting, written procedure, silence procedure and delegated exercise.", "value_kind": "code", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-032" ] }, { "id": "dr-gov-de-dissent-record-ref", "name": "Dissent record reference", "description": "Identifier-only pointer to the dissent artefact owned by the reasoning area; no dissent content is stored locally.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-024" ] } ], "artifacts": [], "inline_only_rationale": "The participation surface is a roster of references plus a small set of coded assertions; the narrative proceedings that would justify a document already exist as minutes or a proceedings record owned by the convening body and, where reasoning is involved, as the reasoning area's artefacts. Holding these facts inline keeps them queryable and avoids creating a rival account of the sitting that this model has no authority to issue." }, { "id": "dr-gov-fnd-interest-recusal", "name": "Declared interests, conflict determination and recusal", "description": "Records interests declared by participants before or during the decision, who determined whether a declared interest was disqualifying, the recusal or mitigation action taken, and the effect of recusal on the applicable participation threshold.", "source_refs": [ "SRC-030", "SRC-032", "SRC-033" ], "questions": [ { "id": "dr-gov-q-coi-declaration", "text": "Which interests were declared by participants in relation to this decision, and against which register entry was each declaration recorded?", "kind": "constraint", "answer_data": [ "Interest declaration entries with declaring party, interest type and declaration instant", "Reference to the interests register entry where one is maintained", "Indicator of whether the declaration was standing or decision-specific" ] }, { "id": "dr-gov-q-coi-determination", "text": "Who determined that a declared interest was disqualifying, and what recusal or mitigation action followed?", "kind": "decision", "answer_data": [ "Determining party reference and determination instant", "Determination outcome code and stated grounds reference", "Recusal or mitigation action code such as withdrawal, non-participation in the vote, or supervised participation" ] }, { "id": "dr-gov-q-coi-threshold-effect", "text": "How does a recusal change the participation threshold or decision rule applied to the affected item?", "kind": "constraint", "answer_data": [ "Adjusted threshold figures after recusal", "Statement of whether the recused participant counted towards presence", "Reference to the provision governing the adjustment" ] }, { "id": "dr-gov-q-coi-retention", "text": "For how long must interest declarations and conflict determinations remain retrievable after the decision is closed, and under what lawful basis?", "kind": "retention", "answer_data": [ "Retention period reference for declarations, separate from the decision's own period", "Lawful basis reference where the declarations contain personal data", "Masking or minimisation requirement applicable at disclosure" ] } ], "data_elements": [ { "id": "dr-gov-de-interest-declaration", "name": "Interest declaration entry", "description": "A declared interest with declaring party reference, interest type code, declaration instant and optional register entry reference.", "value_kind": "object", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-032", "SRC-033" ] }, { "id": "dr-gov-de-conflict-determination", "name": "Conflict determination", "description": "The determination on a declared interest, naming the determining party, outcome code, grounds reference and determination instant.", "value_kind": "object", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-030" ] }, { "id": "dr-gov-de-recusal-action-code", "name": "Recusal action code", "description": "Coded action taken in consequence of a disqualifying interest, such as withdrawal, abstention from the vote or supervised participation.", "value_kind": "code", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-032" ] }, { "id": "dr-gov-de-recusal-effective-time", "name": "Recusal effective time", "description": "Instant from which the recusal took effect, as an RFC 3339 timestamp with explicit offset or Z.", "value_kind": "timestamp", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-028" ] }, { "id": "dr-gov-de-declaration-personal-data-flag", "name": "Declaration personal-data flag", "description": "Indicator that a declaration contains personal data and is therefore subject to minimisation and to a lawful-basis-bound retention period.", "value_kind": "boolean", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-033" ] } ], "artifacts": [], "inline_only_rationale": "Declarations and determinations are short, strongly typed assertions that must remain individually queryable for threshold recomputation and disclosure masking, and they frequently contain personal data whose minimisation is easier to enforce at field level than in a document body. Where a jurisdiction requires a standing register of interests, that register is an external instrument owned by the integrity authority and is referenced rather than reissued here." } ] } ] }, { "id": "dr-gov-bun-lifecycle", "name": "Approval, status lifecycle and change control", "description": "How the decision moved from submission through review to approval or ratification, which validity state it holds and why, how it may be reviewed, appealed, reconsidered or reopened, and how its versions, changes and content provenance are controlled.", "rationale": "Authoritative decision records require an explicit, transition-controlled state model with recorded effective dating, a supersession and consolidation chain, and attributable change history. RFC 2026 demonstrates staged maturity, retirement to a historic state and a bounded appeals procedure; the ELI ontology supplies in-force, consolidation and repeal relations; Akoma Ntoso supplies FRBR-level version identity and lifecycle events; PROV-O supplies attribution, revision and invalidation.", "source_refs": [ "SRC-014", "SRC-012", "SRC-029", "SRC-031" ], "layers": [ { "id": "dr-gov-lay-approval-status", "name": "Submission, approval, status and reconsideration", "description": "The approval pipeline binding, the controlled status and validity model with its effective dating and conditions, and the triggers and windows for review, appeal, reconsideration and reopening.", "source_refs": [ "SRC-026", "SRC-029", "SRC-031" ], "findings": [ { "id": "dr-gov-fnd-approval-binding", "name": "Submission, review and approval stage binding", "description": "Records which governance stages the decision passed, which are mandatory before it can take effect, and the identifier of the external approval or ratification activity that evidences each stage. Distinguishes approval from ratification and from non-binding endorsement.", "source_refs": [ "SRC-012", "SRC-029", "SRC-032" ], "questions": [ { "id": "dr-gov-q-appr-stages", "text": "Which submission, review, approval and ratification stages did this decision pass, and which of them are mandatory before it can take effect?", "kind": "process", "answer_data": [ "Ordered stage entries with stage code, outcome code and completion instant", "Mandatory-stage flag per stage", "Reference to the procedure that prescribes the stage sequence" ] }, { "id": "dr-gov-q-appr-activity-binding", "text": "Which external approval activity record evidences each stage, and which identifier binds the two?", "kind": "interoperability", "answer_data": [ "Approval activity reference per stage resolved against WM-ACT-024", "Binding identifier and the resolution method for it", "Statement that activity execution and completion semantics are not held locally" ] }, { "id": "dr-gov-q-appr-definition", "text": "What distinguishes approval from ratification and from non-binding endorsement in this governance context?", "kind": "definition", "answer_data": [ "Definitions of approval, ratification and endorsement as used by the adopting Dimension", "Legal or procedural effect of each on the decision's validity", "Which of them can retrospectively cure an act taken beyond delegated authority" ] }, { "id": "dr-gov-q-appr-exception", "text": "What happens to the decision when a mandatory stage is skipped, completed out of order, or completed by an unauthorised party?", "kind": "exception", "answer_data": [ "Defect code and the status assigned to a defectively approved decision", "Remediation path such as ratification, re-submission or withdrawal", "Notification obligation and its addressee" ] } ], "data_elements": [ { "id": "dr-gov-de-approval-stage-entry", "name": "Approval stage entry", "description": "One governance stage with stage code, mandatory flag, outcome code, deciding party reference and completion instant.", "value_kind": "object", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-012", "SRC-029" ] }, { "id": "dr-gov-de-approval-activity-ref", "name": "Approval activity reference", "description": "Identifier-only reference to the external approval or ratification activity record that evidences a stage.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-025" ] }, { "id": "dr-gov-de-approval-defect-code", "name": "Approval defect code", "description": "Coded defect in the approval pipeline, such as a skipped mandatory stage or an unauthorised approver.", "value_kind": "code", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-029" ] }, { "id": "dr-gov-de-ratification-ref", "name": "Ratification reference", "description": "Reference to a ratifying act that cures a defect or confirms an act taken under delegated or urgent procedure.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-032" ] } ], "artifacts": [], "inline_only_rationale": "Stages are carried as thin bindings to approval activity records that live in WM-ACT-024; the substantive approval evidence, including approver identity, execution and completion, is that model's artefact. Creating a governance document for the pipeline here would reproduce an approval record this model does not own and would introduce a second, divergent statement of whether approval occurred." }, { "id": "dr-gov-fnd-status-validity", "name": "Status, validity and conditional effectiveness", "description": "Holds the controlled status of the decision and its permitted transitions, the separation of approval, publication and effective instants, the conditions attached to a conditional decision and who confirms them, and identifier-only references to follow-up implementation or monitoring obligations.", "source_refs": [ "SRC-026", "SRC-029", "SRC-031" ], "questions": [ { "id": "dr-gov-q-status-current", "text": "Which status does this decision hold now, and which controlled vocabulary defines the permitted values and the legal transitions between them?", "kind": "state", "answer_data": [ "Current status code from the governed status vocabulary covering draft, submitted, approved, effective, conditional, suspended, superseded, revoked, expired and reopened", "Transition table naming permitted source and target states", "Authority competent to perform each transition" ] }, { "id": "dr-gov-q-status-effective-dating", "text": "From which instant is the decision effective, and how does that differ from the instants at which it was approved and published?", "kind": "temporal", "answer_data": [ "Effective instant, approval instant and publication instant as separate RFC 3339 timestamps with explicit offsets or Z", "Indicator of retrospective effect where the effective instant precedes approval", "Ingestion instant recorded separately from all three" ] }, { "id": "dr-gov-q-status-conditions", "text": "Which conditions precedent or subsequent must hold for a conditional decision to become or remain effective, and who confirms their satisfaction?", "kind": "constraint", "answer_data": [ "Condition entries with condition text reference, type and due instant", "Confirming party reference and confirmation instant", "Consequence code applied when a condition fails" ] }, { "id": "dr-gov-q-status-followup", "text": "Which follow-up implementation or monitoring obligations are attached by reference, and which model owns their execution?", "kind": "relationship", "answer_data": [ "Follow-up obligation references with obligation type and owning model identifier", "Reporting-back instant or interval where one is stipulated", "Explicit statement that implementation progress and outcome measurement are not stored here" ] }, { "id": "dr-gov-q-status-expiry", "text": "What causes automatic expiry as opposed to explicit revocation or suspension, and how is each distinguished in the record?", "kind": "lifecycle", "answer_data": [ "Expiry rule reference and the computed or stipulated expiry instant", "Revocation and suspension entries with acting authority, ground code and effective instant", "Rule preventing an expired decision from being silently treated as revoked" ] } ], "data_elements": [ { "id": "dr-gov-de-status-code", "name": "Status code", "description": "Current governed status of the decision drawn from the controlled status vocabulary.", "value_kind": "code", "cardinality": "1", "required": true, "source_refs": [ "SRC-026", "SRC-031" ] }, { "id": "dr-gov-de-status-transition-entry", "name": "Status transition entry", "description": "A recorded transition with source status, target status, acting authority reference, ground code and effective instant.", "value_kind": "object", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-029", "SRC-031" ] }, { "id": "dr-gov-de-effective-time", "name": "Effective time", "description": "Instant from which the decision has effect, as an RFC 3339 timestamp with explicit offset or Z, held separately from approval and publication instants.", "value_kind": "timestamp", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-028", "SRC-031" ] }, { "id": "dr-gov-de-condition-entry", "name": "Effectiveness condition entry", "description": "A condition precedent or subsequent with type, due instant, confirming party reference and satisfaction state.", "value_kind": "object", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-026" ] }, { "id": "dr-gov-de-followup-obligation-ref", "name": "Follow-up obligation reference", "description": "Identifier-only reference to an implementation or monitoring obligation arising from the decision, together with the model that owns its execution.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-024", "SRC-025" ] } ], "artifacts": [], "inline_only_rationale": "Status, effective dating and conditions are the queryable control fields that determine whether the decision may be relied upon at a given instant, so they must be first-class inline data rather than text inside a document. Follow-up obligations are deliberately reference-only because implementation progress, monitoring results and outcome measurement are executed and evidenced by other models; issuing an artefact here would invite those execution facts to accumulate locally." }, { "id": "dr-gov-fnd-review-appeal-reopening", "name": "Review, appeal, reconsideration and reopening", "description": "Defines what triggers a scheduled or event-driven review, who may appeal or seek reconsideration to which forum within what time limit, on what grounds a closed decision may be reopened and what status it then takes, and how a review outcome is linked to the original decision without overwriting it.", "source_refs": [ "SRC-029", "SRC-030", "SRC-031" ], "questions": [ { "id": "dr-gov-q-rev-trigger", "text": "What triggers a scheduled or event-driven review of this decision, and by which due instant must the review occur?", "kind": "event", "answer_data": [ "Review trigger code covering fixed interval, condition breach, external change and complaint", "Review due instant as an RFC 3339 timestamp with explicit offset or Z", "Party obliged to initiate the review" ] }, { "id": "dr-gov-q-rev-appeal-forum", "text": "Who may appeal or request reconsideration, to which forum, and within what time limit from which starting instant?", "kind": "authority", "answer_data": [ "Eligible appellant classes and the forum reference for each escalation tier", "Time limit expressed as a duration with the instant from which it runs", "Whether the forum's determination is final on procedure, on merits, or on both" ] }, { "id": "dr-gov-q-rev-reopen-grounds", "text": "On what grounds may a closed decision be reopened, and which status does it hold while reopened?", "kind": "exception", "answer_data": [ "Reopening ground codes such as new evidence, procedural defect or material change of circumstance", "Status held during reopening and whether the original remains effective meanwhile", "Authority competent to authorise the reopening" ] }, { "id": "dr-gov-q-rev-outcome-link", "text": "How is the outcome of a review, appeal or reconsideration linked to the original decision without overwriting it?", "kind": "relationship", "answer_data": [ "Outcome reference with relation type such as confirms, varies, remits or quashes", "Reference to an external legal judgment where the forum is judicial", "Rule that the original version remains retrievable in its approved form" ] } ], "data_elements": [ { "id": "dr-gov-de-review-trigger-code", "name": "Review trigger code", "description": "Coded reason a review becomes due, whether periodic, conditional or event-driven.", "value_kind": "code", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-029" ] }, { "id": "dr-gov-de-review-due-time", "name": "Review due time", "description": "Instant by which a due review must be conducted, as an RFC 3339 timestamp with explicit offset or Z.", "value_kind": "timestamp", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-028" ] }, { "id": "dr-gov-de-appeal-window", "name": "Appeal window", "description": "Duration within which an appeal or reconsideration request must be lodged, with the instant from which it runs.", "value_kind": "duration", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-029" ] }, { "id": "dr-gov-de-appeal-forum-ref", "name": "Appeal forum reference", "description": "Reference to the forum competent to hear an appeal at a given escalation tier.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-029", "SRC-030" ] }, { "id": "dr-gov-de-review-outcome-ref", "name": "Review outcome reference", "description": "Reference to the reviewing or appellate determination, with the relation it bears to the original decision.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-030", "SRC-031" ] } ], "artifacts": [], "inline_only_rationale": "Triggers, windows, forums and outcome links are scheduling and relationship data that must drive obligations and queries rather than be read out of a document. The substantive review or appellate determination is itself a decision or a legal judgment owned by another record, so this finding deliberately holds only the pointer and the relation type and creates no artefact that could be mistaken for the determination." } ] }, { "id": "dr-gov-lay-version-provenance", "name": "Version history and provenance of decision content", "description": "The authoritative version chain with supersession and consolidation, the append-only ledger of governance changes with attributable agents and separated time axes, and the provenance of the governed decision content itself.", "source_refs": [ "SRC-014", "SRC-012", "SRC-031" ], "findings": [ { "id": "dr-gov-fnd-change-ledger", "name": "Version identity and append-only change ledger", "description": "Establishes which version is authoritative, which prior versions it supersedes or consolidates, and an append-only ledger of governance changes recording the acting agent, the reason, the authority relied on, the significance of the change and the separated event, observation and ingestion instants.", "source_refs": [ "SRC-014", "SRC-012", "SRC-025", "SRC-028", "SRC-031" ], "questions": [ { "id": "dr-gov-q-ver-authoritative", "text": "Which version of this decision is authoritative now, and which prior versions does it supersede, consolidate or repeal?", "kind": "identity", "answer_data": [ "Version identifier distinct from the decision master identifier", "Supersedes, consolidates and repeals references to prior versions", "Statement of which version applies to a stated instant" ] }, { "id": "dr-gov-q-ver-agent", "text": "For each ledger entry, which agent made the change, for what stated reason, and under which authority?", "kind": "provenance", "answer_data": [ "Acting agent reference with qualified role", "Change reason code and free-reference to the authorising instrument", "Whether the agent acted on delegated authority and under which delegation entry" ] }, { "id": "dr-gov-q-ver-time-axes", "text": "How are event time, observation time and ingestion time recorded separately for a single ledger entry?", "kind": "temporal", "answer_data": [ "Event instant of the governance act itself", "Observation or attestation instant when the fact was witnessed or reported", "Ingestion instant when the Vercy record captured it, each an RFC 3339 timestamp with explicit offset or Z" ] }, { "id": "dr-gov-q-ver-significance", "text": "Which changes are editorial and which are substantive enough to require re-approval or a new version?", "kind": "classification", "answer_data": [ "Change significance code distinguishing editorial, corrective and substantive", "Re-approval requirement flag driven by the significance code", "Erratum reference where an approved version is corrected rather than replaced" ] } ], "data_elements": [ { "id": "dr-gov-de-version-identifier", "name": "Version identifier", "description": "Identifier of a specific governed version of the decision, distinct from and subordinate to the decision master identifier.", "value_kind": "identifier", "cardinality": "1", "required": true, "source_refs": [ "SRC-012", "SRC-031" ] }, { "id": "dr-gov-de-supersedes-ref", "name": "Supersession reference", "description": "Reference from the current version to the version or decision it supersedes, consolidates or repeals, with the relation type.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-031" ] }, { "id": "dr-gov-de-ledger-entry", "name": "Lifecycle ledger entry", "description": "An append-only entry recording one governance act with acting agent, reason code, authority reference, significance code and its separated time axes.", "value_kind": "object", "cardinality": "1..n", "required": true, "source_refs": [ "SRC-014", "SRC-025" ] }, { "id": "dr-gov-de-event-time", "name": "Event time", "description": "Instant at which the recorded governance act occurred, as an RFC 3339 timestamp with explicit offset or Z.", "value_kind": "timestamp", "cardinality": "1", "required": true, "source_refs": [ "SRC-028" ] }, { "id": "dr-gov-de-ingestion-time", "name": "Ingestion time", "description": "Instant at which the Vercy record captured the fact, stored separately from event and observation instants.", "value_kind": "timestamp", "cardinality": "1", "required": true, "source_refs": [ "SRC-028" ] }, { "id": "dr-gov-de-change-significance-code", "name": "Change significance code", "description": "Code distinguishing editorial, corrective and substantive changes and driving the re-approval requirement.", "value_kind": "code", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-029" ] } ], "artifacts": [ { "id": "dr-gov-art-lifecycle-ledger", "name": "Decision lifecycle and change ledger", "description": "An append-only, per-decision ledger of governance acts: status transitions, approvals bound, delegations exercised, versions issued, supersessions, holds placed and released, and disclosure decisions. Entries are never edited or removed; corrections are added as new entries that reference the entry they correct.", "media_or_form": [ "append-only entry sequence", "tabular ledger rendition", "sealed periodic extract" ], "serial": true, "identity_strategy": "Each ledger instance is named {decision-master-identifier}-GOVLED-{NNNN}, where the decision master identifier is the authoritative master-system identifier of the decision, GOVLED is the fixed artefact-kind discriminator, and NNNN is a zero-padded monotonic sequence scoped to that decision master record. Sequence values are never reused and never derived from any date, timestamp or period label; individual entries carry their own ordinal within the instance.", "source_refs": [ "SRC-014", "SRC-025", "SRC-028" ] } ], "inline_only_rationale": null }, { "id": "dr-gov-fnd-content-provenance", "name": "Provenance of the governed decision content", "description": "Records which agents generated, revised and attested the governed content in which qualified roles, from which prior drafts or decisions it was derived, how cited claim and evidence records are referenced without restatement, and whether the record was created natively or imported.", "source_refs": [ "SRC-014", "SRC-012", "SRC-025" ], "questions": [ { "id": "dr-gov-q-prov-attribution", "text": "Which agents generated, revised and attested the governed decision content, and in which qualified roles did they act?", "kind": "provenance", "answer_data": [ "Attribution entries with agent reference and qualified role", "Distinction between drafting, reviewing and attesting acts", "Whether any generating agent was automated and, if so, its identifier" ] }, { "id": "dr-gov-q-prov-derivation", "text": "From which prior decisions, drafts, submissions or precedents was this decision derived?", "kind": "composition", "answer_data": [ "Derivation references with relation type such as revision of, derived from or based on precedent", "Ordering of derivation steps where a chain exists", "Indicator of whether a derivation crosses an organisational boundary" ] }, { "id": "dr-gov-q-prov-citation", "text": "How are cited claim and evidence records referenced so that their content is not restated in this model?", "kind": "interoperability", "answer_data": [ "Claim and evidence references as identifiers with their owning model identifiers", "Resolution method for those identifiers", "Explicit prohibition on copying claim text, evidence appraisal or quality grades locally" ] }, { "id": "dr-gov-q-prov-capture-mode", "text": "How is an imported or migrated decision distinguished from one created natively in this Dimension?", "kind": "identity", "answer_data": [ "Capture mode code covering native creation, import, migration and reconstruction", "Source system reference and the ingestion instant", "Confidence or completeness flag for reconstructed records" ] } ], "data_elements": [ { "id": "dr-gov-de-attribution-entry", "name": "Attribution entry", "description": "An agent's qualified association with the generation, revision or attestation of the governed content.", "value_kind": "object", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-014" ] }, { "id": "dr-gov-de-derivation-ref", "name": "Derivation reference", "description": "Reference to a prior decision, draft or precedent from which this decision was derived, with the relation type.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-014" ] }, { "id": "dr-gov-de-claim-ref", "name": "Cited claim reference", "description": "Identifier-only reference to a claim record owned by the parent knowledge-assertion model.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-024" ] }, { "id": "dr-gov-de-evidence-ref", "name": "Cited evidence reference", "description": "Identifier-only reference to an evidence record; evidence appraisal, quality and custody semantics remain in the evidence model.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-024", "SRC-014" ] }, { "id": "dr-gov-de-capture-mode-code", "name": "Capture mode code", "description": "Code distinguishing native creation, import, migration and reconstruction, with the source system reference.", "value_kind": "code", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-025" ] } ], "artifacts": [], "inline_only_rationale": "Provenance here is a graph of typed references between agents, activities and prior records, and is most defensible when held as data that can be projected into an interoperable provenance vocabulary. The narrative products it points at — the statement of reasons, the reasoning trace and the evidence records — are artefacts already owned by the reasoning area and the evidence model, so no new document is warranted." } ] } ] }, { "id": "dr-gov-bun-stewardship", "name": "Access, integrity and defensible disposition", "description": "How the decision and its governance metadata are classified for access and disclosure, how their integrity is anchored and linked to external audit trails, and how they are retained, held and defensibly disposed of.", "rationale": "Records remain authoritative only if their integrity is demonstrable, their accessibility is controlled and their disposition is defensible. ISO 15489-1 makes integrity, usability and appraisal-driven disposition constitutive of an authoritative record; NARA organises requirements around maintenance, disposal and transfer; NIST separates access enforcement, audit and disposal into controls owned by the operating system; Regulation 1049/2001 and the GDPR supply the disclosure and erasure profiles.", "source_refs": [ "SRC-024", "SRC-025", "SRC-027", "SRC-030", "SRC-033" ], "layers": [ { "id": "dr-gov-lay-access-integrity", "name": "Access classification, disclosure and integrity assurance", "description": "The classification and disclosure schedule for each part of the decision, the exemptions and partial-access rules that apply, and the fixity, sealing and audit-linkage facts that make the record's integrity demonstrable.", "source_refs": [ "SRC-027", "SRC-030", "SRC-033" ], "findings": [ { "id": "dr-gov-fnd-access-disclosure", "name": "Access classification, exemptions and disclosure schedule", "description": "Sets the access class for the decision and, separately, for its rationale and governance metadata; records the exemption grounds relied on to withhold, whether partial access by redaction is possible, when an embargo lapses, what register-level metadata stays discoverable, and which personal data must be minimised before disclosure.", "source_refs": [ "SRC-027", "SRC-030", "SRC-033" ], "questions": [ { "id": "dr-gov-q-acc-class", "text": "What access classification applies, and does it differ between the decision outcome, the rationale and the governance metadata?", "kind": "access", "answer_data": [ "Access class code per addressed part of the record", "Reference to the classification scheme and the classifying authority", "Instant from which each classification applies" ] }, { "id": "dr-gov-q-acc-exemption", "text": "Which exemption grounds justify withholding, and can partial access be granted by releasing the parts not covered by them?", "kind": "exception", "answer_data": [ "Exemption ground codes with the provision reference relied on", "Partial-access feasibility flag and the redaction profile reference", "Whether an overriding public-interest test was applied and its outcome" ] }, { "id": "dr-gov-q-acc-embargo", "text": "When does an embargo lapse so that the decision becomes proactively publishable?", "kind": "temporal", "answer_data": [ "Embargo end instant as an RFC 3339 timestamp with explicit offset or Z", "Publication channel reference and the party obliged to publish", "Behaviour when the embargo lapses while a hold or an appeal is pending" ] }, { "id": "dr-gov-q-acc-discoverable", "text": "Which register-level metadata remains discoverable even when the content itself is withheld?", "kind": "classification", "answer_data": [ "Enumerated discoverable metadata set such as identifier, subject, deciding authority and date", "Rule preventing a reference from itself revealing exempt content", "Handling rule for sensitive items whose very existence is restricted" ] }, { "id": "dr-gov-q-acc-personal-data", "text": "Which personal data in participant and interest records must be minimised or masked before disclosure?", "kind": "privacy", "answer_data": [ "Fields flagged as personal data with their masking rule", "Lawful basis reference for retaining them despite minimisation", "Party accountable for applying the masking before release" ] } ], "data_elements": [ { "id": "dr-gov-de-access-class-code", "name": "Access class code", "description": "Coded access classification applied to an addressed part of the decision record, with the classifying authority reference.", "value_kind": "code", "cardinality": "1..n", "required": true, "source_refs": [ "SRC-027", "SRC-030" ] }, { "id": "dr-gov-de-exemption-ground-code", "name": "Exemption ground code", "description": "Coded ground relied on to withhold, referencing the provision that supplies it.", "value_kind": "code", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-030" ] }, { "id": "dr-gov-de-redaction-profile-ref", "name": "Redaction profile reference", "description": "Reference to the profile describing which parts are removed for a partial release and by what rule.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-030" ] }, { "id": "dr-gov-de-embargo-end-time", "name": "Embargo end time", "description": "Instant at which an embargo lapses, as an RFC 3339 timestamp with explicit offset or Z.", "value_kind": "timestamp", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-028", "SRC-030" ] }, { "id": "dr-gov-de-discoverable-metadata-set", "name": "Discoverable metadata set", "description": "The enumerated set of register-level fields that remain discoverable when content is withheld.", "value_kind": "collection", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-030" ] } ], "artifacts": [ { "id": "dr-gov-art-access-schedule", "name": "Decision access and disclosure schedule", "description": "A per-decision schedule stating the access class of each addressed part, the exemption grounds and provisions relied on, the redaction profile for partial release, embargo instants and the discoverable register metadata. It is a governance instruction, not an access-control implementation or a log of accesses granted.", "media_or_form": [ "structured schedule entries", "tabular disclosure matrix", "publishable register extract" ], "serial": true, "identity_strategy": "Named {decision-master-identifier}-GOVACC-{NNNN}, where the decision master identifier is the authoritative master-system identifier of the decision, GOVACC is the fixed artefact-kind discriminator, and NNNN is a zero-padded monotonic sequence scoped to that decision master record and incremented on each reissue of the schedule. Sequence values are never reused and are never derived from dates, review cycles or period labels.", "source_refs": [ "SRC-025", "SRC-027", "SRC-030" ] } ], "inline_only_rationale": null }, { "id": "dr-gov-fnd-integrity-anchoring", "name": "Fixity, sealing and audit linkage", "description": "Records the digest algorithm and value sealing each governed version, who computed it, which signature, seal or timestamp-token reference evidences non-repudiation of the approval, how a failed verification is reported without altering the sealed version, and which external system holds the audit trail.", "source_refs": [ "SRC-024", "SRC-025", "SRC-027" ], "questions": [ { "id": "dr-gov-q-int-digest", "text": "Which digest algorithm and value seal each governed version, and which agent computed them at which instant?", "kind": "security", "answer_data": [ "Named digest algorithm and its digest value per version", "Computing agent reference and computation instant", "Canonical form over which the digest was computed" ] }, { "id": "dr-gov-q-int-nonrepudiation", "text": "Which signature, seal or timestamp-token reference evidences non-repudiation of the approval?", "kind": "evidence", "answer_data": [ "Signature or seal reference with the signing agent and scheme identifier", "Timestamp-token reference where third-party time attestation exists", "Validation status of the signature at the last verification" ] }, { "id": "dr-gov-q-int-failure", "text": "How is a failed integrity verification reported without altering the sealed version?", "kind": "validation", "answer_data": [ "Verification outcome entry with expected and observed digest values and the verification instant", "Escalation target and notification obligation on failure", "Rule that the sealed version is quarantined rather than corrected in place" ] }, { "id": "dr-gov-q-int-audit-link", "text": "Which external audit or event-log system holds the trail for this record, and which identifier links the two?", "kind": "interoperability", "answer_data": [ "Audit system reference and the correlation identifier", "Statement that audit-record generation, protection and retention are owned by that system", "Coverage note stating which governance acts are expected to appear in the external trail" ] } ], "data_elements": [ { "id": "dr-gov-de-digest-algorithm", "name": "Digest algorithm", "description": "Named cryptographic digest algorithm used to seal a governed version.", "value_kind": "code", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-027" ] }, { "id": "dr-gov-de-digest-value", "name": "Digest value", "description": "Digest computed over the declared canonical form of a governed version.", "value_kind": "text", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-027" ] }, { "id": "dr-gov-de-signature-ref", "name": "Signature or seal reference", "description": "Reference to the signature, seal or timestamp token evidencing non-repudiation of an approval.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-026", "SRC-027" ] }, { "id": "dr-gov-de-verification-outcome-entry", "name": "Verification outcome entry", "description": "Result of an integrity verification with expected and observed values, outcome code and verification instant.", "value_kind": "object", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-027" ] }, { "id": "dr-gov-de-audit-trail-ref", "name": "Audit trail reference", "description": "Correlation reference to the external system that generates and retains the audit trail for this record.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-027" ] } ], "artifacts": [], "inline_only_rationale": "Integrity facts are small, machine-checkable values attached to versions and must be verifiable independently of any document rendition; a document would itself need sealing and would create a recursion. Crucially, the audit trail that these values correlate to is generated, protected and retained by an external system, so this finding holds references only and does not materialise an audit artefact whose semantics this model does not own." } ] }, { "id": "dr-gov-lay-retention-disposition", "name": "Retention, legal hold and defensible disposition", "description": "The retention schedule binding and its trigger, the legal hold state, the authorisation and evidence for disposition, the surviving tombstone facts and the archival transfer path for records of permanent value.", "source_refs": [ "SRC-024", "SRC-025", "SRC-033" ], "findings": [ { "id": "dr-gov-fnd-retention-disposition", "name": "Retention binding, legal hold and defensible disposition", "description": "Binds the decision record to a retention schedule and its start trigger, records whether a legal hold is in force and by whom, states which facts survive when deletion is authorised and which model executes disposition, evidences destruction or transfer, and identifies records of permanent value destined for archival transfer rather than destruction.", "source_refs": [ "SRC-024", "SRC-025", "SRC-027", "SRC-033" ], "questions": [ { "id": "dr-gov-q-ret-schedule", "text": "Which retention schedule and issuing authority govern this decision record, and which event starts the retention clock?", "kind": "retention", "answer_data": [ "Retention schedule reference with issuing authority and item number", "Retention trigger code such as decision effective, superseded, closed or last accessed", "Retention period expressed as a duration from the trigger instant" ] }, { "id": "dr-gov-q-ret-hold", "text": "Is a legal hold or disposition freeze in force, who imposed it, and when was it released?", "kind": "state", "answer_data": [ "Hold entries with imposing authority reference, ground and scope", "Imposition and release instants as RFC 3339 timestamps with explicit offset or Z", "Rule suspending all disposition actions while any hold is in force" ] }, { "id": "dr-gov-q-ret-tombstone", "text": "When deletion is authorised, which facts survive in the tombstone and which model or Dimension policy executes the disposition?", "kind": "lifecycle", "answer_data": [ "Surviving tombstone fields: decision master identifier, version identifier, deciding authority reference, status at disposition, disposition authorisation reference and disposition instant", "Named executing owner: the carrying record model and the adopting Dimension's records authority", "Statement of which erasure rule, if any, overrides the tombstone-first default" ] }, { "id": "dr-gov-q-ret-evidence", "text": "How is destruction, transfer or migration evidenced so that the disposition remains defensible?", "kind": "evidence", "answer_data": [ "Disposition evidence reference such as a certificate or receipt identifier", "Authorising party reference and the authorisation instant", "Ledger entry recording the disposition act with its separated time axes" ] }, { "id": "dr-gov-q-ret-permanent", "text": "Which decisions are appraised as being of permanent value and transferred to an archival authority rather than destroyed?", "kind": "decision", "answer_data": [ "Appraisal outcome code and the appraising authority reference", "Archival destination reference and transfer instant", "Transfer format and metadata expectations stated by the receiving authority" ] } ], "data_elements": [ { "id": "dr-gov-de-retention-schedule-ref", "name": "Retention schedule reference", "description": "Reference to the retention schedule item and its issuing authority that governs this record.", "value_kind": "reference", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-024", "SRC-025" ] }, { "id": "dr-gov-de-retention-trigger-code", "name": "Retention trigger code", "description": "Coded event that starts the retention period, with the instant at which the trigger occurred.", "value_kind": "code", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-025" ] }, { "id": "dr-gov-de-hold-entry", "name": "Legal hold entry", "description": "A hold on disposition with imposing authority reference, ground, scope, imposition instant and release instant.", "value_kind": "object", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-025", "SRC-033" ] }, { "id": "dr-gov-de-disposition-action-code", "name": "Disposition action code", "description": "Coded disposition outcome such as retain, destroy, transfer, migrate or tombstone.", "value_kind": "code", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-024", "SRC-025" ] }, { "id": "dr-gov-de-tombstone-record", "name": "Tombstone record", "description": "The minimal surviving set of identity, authority, status and disposition facts retained after an authorised deletion of content.", "value_kind": "object", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-025", "SRC-033" ] }, { "id": "dr-gov-de-transfer-destination-ref", "name": "Archival transfer destination reference", "description": "Reference to the archival authority or repository receiving a record appraised as permanent.", "value_kind": "reference", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-025" ] } ], "artifacts": [ { "id": "dr-gov-art-disposition-record", "name": "Decision review, hold and disposition record", "description": "A per-decision record of appraisal outcome, retention binding, holds imposed and released, disposition authorisation and the evidence of destruction, transfer or migration. It states and evidences the disposition decision; it does not perform the disposition, which is executed by the carrying record model and the adopting Dimension's records authority.", "media_or_form": [ "structured disposition entries", "certificate or receipt rendition", "archival transfer manifest extract" ], "serial": true, "identity_strategy": "Named {decision-master-identifier}-GOVDIS-{NNNN}, where the decision master identifier is the authoritative master-system identifier of the decision, GOVDIS is the fixed artefact-kind discriminator, and NNNN is a zero-padded monotonic sequence scoped to that decision master record and incremented for each disposition cycle. Sequence values are never reused and never derived from a date, retention year or schedule period.", "source_refs": [ "SRC-024", "SRC-025", "SRC-033" ] } ], "inline_only_rationale": null } ] } ] } ] }, "functions": [ { "id": "dr-core-fn-frame", "name": "Frame the decision", "description": "Establish the governed decision record: mint or bind its identifier, classify it against a decision type and jurisdictional profile, state the decision question and admissible answer space, record the triggering issue, objectives and critical success factors, and declare context dependencies and binding constraints.", "inputs": [ "Triggering issue or change-request reference", "Decision type code and jurisdictional profile identifier", "Decision question text and admissible answer space", "Objectives, critical success factors and impacted indicator references", "Declared constraints, required input references and upstream decision references" ], "outputs": [ "Framed decision record carrying a resolvable identifier and version discriminator", "Recorded decision question with its admissible answer space", "Objective, constraint and dependency sets", "Derived list of profile-required content elements" ], "preconditions": [ "An identifier is available from the identity priority order, master-system identifier first", "The decision type and its bound profile are published by the profile custodian" ], "effects": [ "A uniquely identified decision record exists and is resolvable", "Profile-required content elements are declared so later completeness checks are decidable", "Framing content becomes the fixed reference against which the outcome statement is validated" ], "source_refs": [ "SRC-001", "SRC-005", "SRC-006", "SRC-008" ] }, { "id": "dr-core-fn-register-alternative", "name": "Register a candidate alternative", "description": "Add a candidate alternative to the decision's alternative set with its own identifier, kind code and description, including registration of a no-action or business-as-usual baseline where the bound profile requires one.", "inputs": [ "Decision identifier", "Alternative description sufficient for evaluation", "Alternative kind code, including baseline or do-minimum where applicable", "Variant-of reference where the alternative varies another" ], "outputs": [ "Registered alternative with its own identifier and registration ordinal", "Updated alternative set for the decision", "Baseline presence indicator for the bound profile" ], "preconditions": [ "The decision has been framed and its profile binding is known", "The alternative is described in terms that the declared criteria can be applied to" ], "effects": [ "The alternative set is extended append-only; registered alternatives are never silently removed", "Absence of a profile-required baseline becomes a detectable validation failure", "Variants remain distinguishable from independently registered alternatives" ], "source_refs": [ "SRC-006", "SRC-009", "SRC-008" ] }, { "id": "dr-core-fn-declare-criteria", "name": "Declare the evaluation criteria and weights", "description": "Record the criteria against which alternatives will be compared, each with its measure, scale, unit and preference direction, the separation of mandatory pass-or-fail criteria from compensatory ones, and the weighting scheme with its declared method or an explicit statement that no weighting is used.", "inputs": [ "Objectives and critical success factors from the framed decision", "Candidate criteria with measures, scales and preference directions", "Weighting method identifier and weight values where a weighting scheme applies", "Monetisation convention where a criterion is expressed in monetary terms" ], "outputs": [ "Declared criterion set addressable by criterion identifier", "Declared weighting scheme, or an explicit no-weighting statement", "Classification of criteria as mandatory or compensatory" ], "preconditions": [ "Objectives and constraints have been framed", "Each criterion is stated in terms that permit a result to be recorded against it" ], "effects": [ "Alternatives can be scored against a fixed and citable criterion set", "Criteria that also act as eligibility filters are linked to the screening criteria", "Weighting remains optional so that decisions without a weighting scheme are still complete" ], "source_refs": [ "SRC-007", "SRC-008", "SRC-006" ] }, { "id": "dr-core-fn-evaluate", "name": "Evaluate and compare the alternatives", "description": "Record, as governed decision content, one evaluation round: eligibility screening results with reasons for exclusion, per-criterion results for admitted alternatives, any aggregation and ranking, sensitivity findings, and citations to evidence. Running an executable rule set or decision service to produce those results is outside this model.", "inputs": [ "Declared criterion set and screening criteria", "Registered alternative set", "Recorded results per alternative and criterion", "Evidence item references supporting results or assumptions" ], "outputs": [ "Eligibility and exclusion results with a recorded reason for each exclusion", "Alternative-by-criterion comparison matrix for the round", "Aggregation, ranking and sensitivity findings", "Result status markers for contested, provisional or bias-adjusted results" ], "preconditions": [ "At least one alternative is registered and the criterion set is declared", "The evaluation round ordinal is available and has not been used" ], "effects": [ "A serially numbered, append-only evaluation round is recorded and never edited in place", "Every excluded alternative carries an explicit recorded reason", "Cited evidence is bound by identifier and role only; evidence appraisal, integrity and provenance stay with the evidence model", "Execution traces of any external decision service are not captured here" ], "source_refs": [ "SRC-007", "SRC-008", "SRC-006", "SRC-009" ] }, { "id": "dr-core-fn-record-outcome", "name": "Record the selected decision outcome", "description": "Record the selected alternative, the disposition, the explicit outcome statement answering the decision question, the declared effect onset and applicability, the factors expressly balanced, any explained divergence from the ranking, and the attached conditions, unresolved matters and stated consequences.", "inputs": [ "Selected alternative reference or declared variant description", "Disposition code from the controlled list", "Outcome statement text", "Balanced factors and divergence explanation where the top-ranked alternative was not selected", "Attached conditions, unresolved matters, accepted risks and reopening triggers" ], "outputs": [ "Outcome statement bound to the recorded decision question", "Disposition code and selected alternative reference", "Declared effect onset and applicability", "Condition, reservation, consequence and raised-concern sets" ], "preconditions": [ "The alternative set contains the selected alternative or the alternative the selected variant derives from", "Profile-required comparison and baseline content is present", "The outcome statement falls within the recorded admissible answer space where one is declared" ], "effects": [ "The decision question has an explicit, immutable recorded answer; changing it requires a superseding decision that cites this one", "Divergence from the evaluation ranking carries a recorded explanation where it occurs", "Approval execution, publication, filing and enforcement of the outcome remain with the referenced activity, record and implementation models", "Any monitoring or compliance work the conditions direct is executed and tracked by those models, not here" ], "source_refs": [ "SRC-010", "SRC-009", "SRC-006", "SRC-001" ] }, { "id": "dr-rsn-fn-bind-reason-grounding", "name": "Bind a reason to governed claims and evidence citations", "description": "Attach claim references and evidence citation bindings to a reason or premise, recording citation type, locator and stance, and reporting any reference that does not resolve in its owning model.", "inputs": [ "reason or premise record identifier", "claim references with version pins", "evidence references with citation type, locator and stance" ], "outputs": [ "persisted binding set attached to the reason", "resolution report listing resolved, superseded and dangling references" ], "preconditions": [ "The reason record exists and names the decision it explains by its master-system identifier", "Citation types are drawn from a governed citation-typing vocabulary with a declared version", "Referenced claim and evidence identifiers are resolvable in WM-KNW-007 and WM-KNW-008 respectively" ], "effects": [ "Creates or revises binding records only; never creates, alters, grades or retracts a referenced claim or evidence item", "Records unresolvable references as declared gaps rather than substituting local content" ], "source_refs": [ "SRC-018", "SRC-023", "SRC-015" ] }, { "id": "dr-rsn-fn-assemble-argument-structure", "name": "Assemble premises, inference and conclusion into an argument structure", "description": "Record an argument as premises, a first-class inference element carrying a scheme reference, and a conclusion, together with asserted support and conflict relations and any critical-question responses.", "inputs": [ "argument element references with role codes", "inference scheme reference and catalogue version", "asserted support and conflict relations", "critical-question responses or unanswered markers" ], "outputs": [ "persisted argument structure record", "argument structure export projection" ], "preconditions": [ "Every premise and conclusion slot references a governed claim or is explicitly marked as a local restatement", "The scheme reference resolves in a declared catalogue, or the argument is flagged as informally schemed" ], "effects": [ "Records asserted inference and conflict only; performs no validity, soundness or entailment evaluation and issues no acceptability verdict", "Leaves premise support status as asserted, without inferring support from the presence of a binding" ], "source_refs": [ "SRC-013", "SRC-015", "SRC-019" ] }, { "id": "dr-rsn-fn-recheck-assumption-and-sensitivity", "name": "Re-check declared assumptions and sensitivity statements", "description": "Compare declared assumptions, dependencies and sensitivity statements against the current state of the inputs and references they name, and annotate each with a status so that stale reasoning is visible.", "inputs": [ "assumption and dependency records with validity windows and defeat conditions", "sensitivity variable references and recorded ranges", "current resolution state of the referenced inputs" ], "outputs": [ "assumption status annotations", "list of stale, defeated-by-declaration or unresolvable references" ], "preconditions": [ "Each assumption declares either a validity window, a defeat condition or an explicit unbounded marker", "The referenced inputs expose a resolvable identity and revision" ], "effects": [ "Annotates existing records with a status and the time of the check; never edits the asserted assumption text", "Does not execute the underlying analysis, does not re-run any model, does not enforce an outcome and does not change any decision state", "Escalation, review scheduling and any consequent decision action are left to the `governance` split and the adopting Dimension" ], "source_refs": [ "SRC-022", "SRC-016", "SRC-019" ] }, { "id": "dr-rsn-fn-record-objection", "name": "Record an objection, counter-evidence or dissenting position", "description": "Capture an objection or counter-argument against a named reason or inference, with its attribution, any counter-evidence binding, and an appended disposition where one has been stated.", "inputs": [ "objection statement and target reference", "objecting party reference and stated capacity", "counter-evidence citation bindings", "disposition statement where available" ], "outputs": [ "persisted objection record with attribution", "dissenting or minority opinion artefact where a position is preserved separately" ], "preconditions": [ "The targeted reason, premise or inference exists", "The objecting party is identified by reference to a record held outside this model" ], "effects": [ "Appends objection and disposition records; existing objections and dissent are never overwritten or removed", "Applies argument-level defeated markings without altering the recorded outcome or any decision state", "Adjudication authority, quorum and approval workflow remain with the `governance` split and are not exercised here" ], "source_refs": [ "SRC-019", "SRC-015", "SRC-013" ] }, { "id": "dr-rsn-fn-render-bounded-explanation", "name": "Produce a bounded, audience-scoped explanation rendition", "description": "Generate an explanation for a declared audience from a named rationale revision, attaching the omission set, the faithfulness assertion and the provenance of any machine-generated text.", "inputs": [ "rationale record set identifier and revision", "declared audience and the entitlement or duty being answered", "omission policy reference from the `governance` split" ], "outputs": [ "explanation rendition artefact with content digest", "omission record stating what was withheld and on what ground" ], "preconditions": [ "The audience is declared and an omission policy reference resolves", "The rationale revision to be rendered is resolvable and its provenance records are complete", "Any machine-generated content carries its generating agent and version" ], "effects": [ "Emits a rendition linked to the rationale revision it renders, labelled where content was machine-generated", "Creates no entitlement, discharges no regulatory duty and makes no representation that the reasoning is true, lawful or optimal", "Is not an audit record: audit trails, access logging and disclosure approval remain with the `governance` split" ], "source_refs": [ "SRC-016", "SRC-017", "SRC-014" ] }, { "id": "dr-gov-fn-register-authority-mandate", "name": "Register deciding authority and mandate binding", "description": "Create or update the register entry that binds a deciding authority to a mandate instrument, its scope limits and its validity window, and attach that binding to a decision.", "inputs": [ "Deciding authority reference", "Mandate instrument reference", "Scope limits and validity window", "Decision master identifier" ], "outputs": [ "Authority register entry identifier", "Authority binding attached to the decision", "Lifecycle ledger entry" ], "preconditions": [ "The authority reference resolves in the party or governing-body register", "The mandate instrument reference resolves and its validity window covers the decision event time", "The caller holds the governance owner or decision authority role" ], "effects": [ "The decision carries a resolvable authority and mandate binding", "Any subsequent authority claim is validated against this register entry", "The registration is recorded as an append-only ledger entry with separated event and ingestion instants" ], "source_refs": [ "SRC-024", "SRC-014", "SRC-031" ] }, { "id": "dr-gov-fn-record-delegation", "name": "Record or revoke a delegation step", "description": "Add a delegation or sub-delegation step with its ceiling, validity window and sub-delegation permission, or revoke an existing step and record the retrospectivity of the revocation.", "inputs": [ "Delegator and delegate references", "Delegation instrument reference", "Ceiling, validity window and sub-delegation flag", "Revocation instrument reference where revoking" ], "outputs": [ "Delegation chain entry or revocation entry", "Updated delegation chain for the decision", "Lifecycle ledger entry" ], "preconditions": [ "The delegator holds a mandate or an upstream delegation that permits sub-delegation", "The delegation window falls inside the upstream mandate window" ], "effects": [ "The competence chain from mandate holder to acting agent becomes traceable", "Revocation marks the step closed without deleting it and records whether prior acts remain valid" ], "source_refs": [ "SRC-014", "SRC-032" ] }, { "id": "dr-gov-fn-declare-interest-recusal", "name": "Declare an interest and record a recusal", "description": "Capture a participant's declared interest, the determination made on it and any resulting recusal or mitigation action, together with its effective instant.", "inputs": [ "Declaring participant reference", "Interest type and declaration instant", "Determining party reference and outcome", "Recusal or mitigation action code" ], "outputs": [ "Interest declaration entry", "Conflict determination entry", "Recusal action with effective instant" ], "preconditions": [ "The participant appears on the decision's participant roster", "The determining party is distinct from the declaring participant" ], "effects": [ "The participation record shows who was excluded and from which instant", "Any recorded participation threshold figures are re-stated to reflect the recusal", "Personal-data flags drive masking on later disclosure" ], "source_refs": [ "SRC-030", "SRC-032", "SRC-033" ] }, { "id": "dr-gov-fn-attest-decision-rule", "name": "Attest satisfaction of the decision rule", "description": "Record the presiding authority's own assertion that the applicable decision rule and participation threshold were satisfied, with the counted and required figures as asserted.", "inputs": [ "Decision rule code and prescribing provision reference", "Counted and required participation figures", "Attesting party reference", "Participation mode code" ], "outputs": [ "Quorum attestation entry", "Participation mode recorded on the decision", "Lifecycle ledger entry" ], "preconditions": [ "A participant roster exists for the decision", "The attesting party holds a presiding or secretariat role for the proceeding" ], "effects": [ "The validity of the proceeding becomes inspectable as an attributed assertion", "No threshold computation, evaluation or enforcement is performed or claimed by this model; the attestation records what the authority asserted" ], "source_refs": [ "SRC-029", "SRC-032" ] }, { "id": "dr-gov-fn-bind-approval-act", "name": "Bind an approval or ratification act", "description": "Attach an identifier-only reference to an external approval, ratification or endorsement activity record for a named governance stage and record the stage outcome.", "inputs": [ "Stage code and mandatory flag", "Approval activity reference in the approval activity model", "Stage outcome code and completion instant" ], "outputs": [ "Approval stage entry", "Updated pipeline completeness state", "Lifecycle ledger entry" ], "preconditions": [ "The referenced approval activity identifier resolves in the approval activity model", "The stage is defined in the applicable procedure" ], "effects": [ "The decision's approval pipeline becomes traceable to external evidence", "Approval execution, assignment and completion semantics remain owned by the referenced activity model" ], "source_refs": [ "SRC-012", "SRC-025", "SRC-032" ] }, { "id": "dr-gov-fn-transition-status", "name": "Apply a governed status transition", "description": "Move the decision from one governed status to another — including to effective, conditional, suspended, superseded, revoked, expired or reopened — recording the acting authority, ground, effective instant and any successor reference.", "inputs": [ "Target status code", "Acting authority reference and ground code", "Effective instant", "Successor or superseding version reference where applicable" ], "outputs": [ "Status transition entry", "Updated current status", "Lifecycle ledger entry" ], "preconditions": [ "The transition is permitted from the current status by the governed transition table", "The acting authority is competent for that transition", "No unreleased legal hold blocks a transition to a disposition-bearing state" ], "effects": [ "The decision's reliability at a given instant becomes determinable", "Superseded versions remain retrievable; no approved version is mutated by a transition" ], "source_refs": [ "SRC-026", "SRC-029", "SRC-031" ] }, { "id": "dr-gov-fn-open-reconsideration", "name": "Open a review, appeal or reopening", "description": "Open a review, appeal or reconsideration against the decision, recording the trigger or appellant, the competent forum, the applicable window and the status held while the matter is open.", "inputs": [ "Trigger or appellant reference", "Forum reference and escalation tier", "Window start instant and duration", "Reopening ground code where reopening" ], "outputs": [ "Review or appeal entry with its due and expiry instants", "Status entry reflecting the open matter", "Lifecycle ledger entry" ], "preconditions": [ "The requesting party falls within an eligible appellant class for the tier", "The request instant falls inside the applicable window" ], "effects": [ "The reconsideration obligation becomes trackable and time-bounded", "The original version remains retrievable in its approved form while the matter is open" ], "source_refs": [ "SRC-029", "SRC-030" ] }, { "id": "dr-gov-fn-classify-and-schedule-disclosure", "name": "Classify access and schedule disclosure", "description": "Set or revise the access class for each addressed part of the decision, record the exemption grounds and redaction profile, and set embargo and proactive-publication instants.", "inputs": [ "Addressed part identifiers", "Access class codes and classifying authority reference", "Exemption ground codes and provision references", "Embargo end instant and publication channel reference" ], "outputs": [ "Access and disclosure schedule version", "Discoverable metadata set", "Lifecycle ledger entry" ], "preconditions": [ "The classifying authority holds the release or classification role", "Any exemption ground cites a resolvable provision reference" ], "effects": [ "Downstream systems obtain a declarative instruction for release and redaction", "Enforcement of the classification is performed by the access-control system, not by this model" ], "source_refs": [ "SRC-027", "SRC-030", "SRC-033" ] }, { "id": "dr-gov-fn-set-hold-state", "name": "Impose or release a legal hold", "description": "Record the imposition or release of a legal hold or disposition freeze over the decision record, with the imposing authority, ground, scope and instants.", "inputs": [ "Imposing or releasing authority reference", "Hold ground and scope", "Imposition or release instant" ], "outputs": [ "Hold entry with updated hold state", "Suspension flag on pending disposition actions", "Lifecycle ledger entry" ], "preconditions": [ "The imposing party holds legal or records authority in the adopting Dimension", "A release names the hold entry it releases" ], "effects": [ "All disposition actions are blocked while any hold is in force", "Hold history remains retrievable after release for defensibility" ], "source_refs": [ "SRC-025", "SRC-033" ] }, { "id": "dr-gov-fn-request-disposition", "name": "Request disposition or tombstoning", "description": "Request a disposition outcome for the decision record — retain, destroy, transfer, migrate or tombstone — with its authorisation, and record the evidence returned by the executing custodian.", "inputs": [ "Retention schedule reference and trigger instant", "Requested disposition action code", "Authorising party reference", "Disposition evidence reference returned by the executor" ], "outputs": [ "Disposition record entry", "Tombstone record where content is removed", "Lifecycle ledger entry" ], "preconditions": [ "The retention period has elapsed from the recorded trigger instant", "No hold is in force", "An appraisal outcome exists for records of potential permanent value" ], "effects": [ "Execution is performed by the carrying record model and the adopting Dimension's records authority, not here", "Identity, authority, status and disposition facts survive in the tombstone unless an authorised erasure rule requires their removal" ], "source_refs": [ "SRC-024", "SRC-025", "SRC-033" ] }, { "id": "dr-gov-fn-seal-and-verify-integrity", "name": "Seal a version and verify its integrity", "description": "Compute and store the digest over the declared canonical form of a governed version, attach any signature or timestamp-token reference, and later verify a retrieved projection against the stored digest.", "inputs": [ "Version identifier and canonical form declaration", "Digest algorithm identifier", "Signature or timestamp-token reference where available", "Retrieved projection for verification" ], "outputs": [ "Digest value bound to the version", "Verification outcome entry", "Lifecycle ledger entry" ], "preconditions": [ "The canonicalization rules for the projection are declared", "The sealing agent is identified" ], "effects": [ "Integrity of a version becomes independently checkable", "A failed verification quarantines the projection and raises a notification; the sealed version is never corrected in place", "Audit-record generation and retention remain with the referenced audit system" ], "source_refs": [ "SRC-024", "SRC-025", "SRC-027" ] } ], "composition": [ { "target": "WM-KNW-007 Claim", "relation": "CHILD", "purpose": "Register the decision under its knowledge parent. Generic assertion identity, truth-value, endorsement and claim-conflict machinery stay with the parent; this model adds only decision-specific specialisation, namely the alternative set, criteria, evaluation and disposition.", "required": true, "source_refs": [ "SRC-003", "SRC-012" ] }, { "target": "WM-KNW-008 Evidence", "relation": "REFERENCE", "purpose": "Bind evidence cited in support of an evaluation result or a declared assumption, carrying the evidence identifier and the role of the citation. Evidence appraisal, sufficiency, integrity and chain of custody remain with the target and are never reproduced here.", "required": false, "source_refs": [ "SRC-008", "SRC-005" ] }, { "target": "WM-ACT-024 Decision-making / approval activity", "relation": "REFERENCE", "purpose": "Reference the activity that produced and approved the decision, so that participants, deliberation, quorum and approval execution can be resolved. This model carries the reference and any decision-specific binding parameters only; the activity's lifecycle and execution semantics stay with the target.", "required": false, "source_refs": [ "SRC-001", "SRC-004" ] }, { "target": "WM-REC-010 Document / record", "relation": "REFERENCE", "purpose": "Reference the record that expresses, files or publishes this decision content, following the Work and Expression separation from Manifestation and Item. Rendering, formatting, publication, preservation and record-level retention execution stay with the target.", "required": false, "source_refs": [ "SRC-012" ] }, { "target": "WM-POL-021 Legal judgment", "relation": "ALIGN", "purpose": "Where the decision is judicial, align the outcome content to the judgment's decision section and its background and motivation structure. The judgment's binding legal force, enforceability and citation regime remain with the target; alignment asserts no conformance.", "required": false, "source_refs": [ "SRC-012", "SRC-005" ] }, { "target": "WM-ACT-032 Change request", "relation": "REFERENCE", "purpose": "Carry the back-reference from a decision to the change request it resolves, matching the incoming relation in the ledger. The request's own lifecycle, prioritisation and the implementation, monitoring and compliance work that follows the decision belong to the target.", "required": false, "source_refs": [ "SRC-010", "SRC-004" ] }, { "target": "OMG Decision Model and Notation (DMN) version 1.5", "relation": "ALIGN", "purpose": "Align framing vocabulary to DMN element names for interchange: question, allowedAnswers, informationRequirement, knowledgeRequirement, supportedObjective, impactedPerformanceIndicator and decision table hitPolicy. Alignment only; no conformance is claimed, DMN identifiers are document-scoped and do not satisfy the identity priority, and execution of DMN decision logic stays with a decision service.", "required": false, "source_refs": [ "SRC-001", "SRC-002" ] }, { "target": "Jurisdictional appraisal and record-of-decision profiles (HM Treasury Green Book, US NEPA record-of-decision pattern)", "relation": "MIX-IN", "purpose": "Bind an optional jurisdictional profile that adds required content elements, such as a mandatory business-as-usual baseline, recorded reasons for eliminating alternatives, identification of a preferable alternative and declared mitigation conditions. Profiles are opt-in per decision type; none is a global default and the NEPA pattern is cited as historical.", "required": false, "source_refs": [ "SRC-006", "SRC-009", "SRC-010", "SRC-011" ] }, { "target": "WM-KNW-007", "relation": "CHILD", "purpose": "Registry-declared parent. Decision rationale is a specialisation of governed assertion content: premises and conclusions are references to claims that WM-KNW-007 identifies, versions and verifies. This model adds only the argument role a claim occupies, the version pinned at reasoning time and any marked local restatement; it never issues claim identity, changes claim wording or records claim verification status.", "required": true, "source_refs": [ "SRC-015", "SRC-013" ] }, { "target": "WM-KNW-008", "relation": "REFERENCE", "purpose": "Decision rationale cites evidence. This model carries the evidence reference, the citation type, the locator into the cited resource, the citation stance and the decision-local appraisal of how far the evidence bears on this decision. Evidence identity, acquisition, chain of custody, intrinsic certainty grading and evidence retention remain in WM-KNW-008 and are carried by reference; a referenced certainty grade is read and reproduced unchanged with its read time, never recomputed here.", "required": true, "source_refs": [ "SRC-018", "SRC-023", "SRC-020" ] }, { "target": "WM-ACT-032", "relation": "REFERENCE", "purpose": "Incoming registry-declared relation: a change request is resolved by a decision and references the decision record. This model exposes stable rationale identifiers so that such a reference can resolve, and nothing more. Change-request intake, triage, workflow and closure are owned by WM-ACT-032, and no function here advances or evaluates a change request.", "required": false, "source_refs": [ "SRC-014" ] }, { "target": "OMG Structured Assurance Case Metamodel (SACM) 2.3, Argumentation package", "relation": "ALIGN", "purpose": "Alignment target for argument structure: claims, argument reasoning, asserted inference, asserted evidence and counter-argument or counter-evidence links. Mapping is asserted at the level of element roles only; no conformance is claimed and the alignment does not import SACM's assurance-case lifecycle or its notation.", "required": false, "source_refs": [ "SRC-015" ] }, { "target": "Argument Interchange Format (AIF) core ontology", "relation": "ALIGN", "purpose": "Alignment target for representing inference, conflict and preference as first-class nodes and for scheme forms carrying premises, conclusions, exceptions and presumptions. Used to justify first-class inference elements and scheme references; AIF's dialogue extensions are outside this model's scope.", "required": false, "source_refs": [ "SRC-013" ] }, { "target": "W3C PROV-O", "relation": "ALIGN", "purpose": "Alignment target for reasoning provenance: attribution of rationale to a person or software agent, association with a generating activity, derivation from prior rationale, and generation time. Adopting the vocabulary does not transfer any responsibility, access or retention semantics to this model.", "required": false, "source_refs": [ "SRC-014" ] }, { "target": "CiTO, the Citation Typing Ontology", "relation": "ALIGN", "purpose": "Alignment target for the citation-type vocabulary used in evidence bindings, including cited as evidence, cited as authority, supports, disagrees with and refutes. Citation types characterise the citing act only and are never interpreted as evidential weight or quality grades.", "required": false, "source_refs": [ "SRC-018" ] }, { "target": "W3C Web Annotation Data Model", "relation": "ALIGN", "purpose": "Alignment target for locators that pinpoint the cited part of a resource using quotation, position or fragment selectors, and for creator, created, generator and generated provenance on annotations. Used to reference precisely without copying cited content into this model.", "required": false, "source_refs": [ "SRC-023" ] }, { "target": "Goal Structuring Notation Community Standard, Version 3", "relation": "ALIGN", "purpose": "Alignment target for separating assumption from supported goal, justification from claim, and solution as a reference to evidence, and for the dialectic notions of challenge and defeated as argument-level annotations. No claim is made that a rationale record constitutes a GSN-conformant assurance case.", "required": false, "source_refs": [ "SRC-019" ] }, { "target": "GRADE certainty-of-evidence framework as rendered in the ACIP GRADE Handbook", "relation": "ALIGN", "purpose": "Alignment target for carrying a referenced certainty grade and for distinguishing the grading of a body of evidence from a decision-local appraisal. Grading domains, methodology and the act of grading remain with the evidence model; this model reproduces the grade with its framework identifier and read time.", "required": false, "source_refs": [ "SRC-020" ] }, { "target": "WM-KNW-007", "relation": "EXTEND", "purpose": "Specialises the parent knowledge-assertion model for the decision subject by adding decision-specific mandate, participation, approval and validity-state governance. Generic assertion identity, claim typing and evidential binding machinery stay in the parent and are not restated here.", "required": true, "source_refs": [ "SRC-024", "SRC-014" ] }, { "target": "WM-KNW-008", "relation": "REFERENCE", "purpose": "Carries identifier-only references from decision content provenance to evidence records. Evidence identity, appraisal, quality grading and chain-of-custody semantics remain in the target; rationale-level evidence binding is owned by the reasoning area of this model.", "required": false, "source_refs": [ "SRC-024", "SRC-014" ] }, { "target": "WM-ACT-024", "relation": "REFERENCE", "purpose": "Binds each approval, ratification or endorsement stage to the external activity record that evidences it. This model stores the binding identifier, stage code and outcome; activity initiation, assignment, execution and completion remain owned by the target.", "required": false, "source_refs": [ "SRC-012", "SRC-025" ] }, { "target": "WM-REC-010", "relation": "REFERENCE", "purpose": "Names the record that carries the decision and the authority that executes retention, transfer and destruction. This model states the retention binding, hold state, disposition authorisation and tombstone content; repository placement and physical execution remain owned by the target.", "required": true, "source_refs": [ "SRC-024", "SRC-025" ] }, { "target": "WM-POL-021", "relation": "REFERENCE", "purpose": "References a judicial or tribunal judgment that determines an appeal against the decision. Court composition, pleadings, orders, remedies and judicial reasoning are not modelled here.", "required": false, "source_refs": [ "SRC-030" ] }, { "target": "WM-ACT-032", "relation": "REFERENCE", "purpose": "Inbound relation: a change request is resolved by a decision. This model retains only the resolved change-request identifier and the resolution instant; change-request intake, triage, states and closure remain owned by the source model.", "required": false, "source_refs": [ "SRC-024", "SRC-025" ] }, { "target": "Policy and executable-rule models (WM-POL family)", "relation": "REFERENCE", "purpose": "A decision may be constrained by, or may establish, a policy or executable rule. Only the reference, its binding role and any decision-specific parameters are held here; rule authoring, runtime evaluation, obligation discharge and enforcement remain external and are never inferred from the reference.", "required": false, "source_refs": [ "SRC-027", "SRC-030" ] }, { "target": "ISO 15489-1:2016 records management concepts and principles", "relation": "ALIGN", "purpose": "Aligns authoritativeness, metadata, assigned responsibilities, records controls and appraisal-driven disposition. Alignment only: no conformance is claimed, since conformance depends on the adopting Dimension's records system and policy.", "required": false, "source_refs": [ "SRC-024" ] }, { "target": "W3C PROV-O provenance ontology", "relation": "ALIGN", "purpose": "Maps attribution, qualified association with roles, delegation, revision, derivation and invalidation onto local provenance and delegation structures for interchange. Local semantics prevail where they diverge.", "required": false, "source_refs": [ "SRC-014" ] }, { "target": "ELI ontology version 1.5", "relation": "ALIGN", "purpose": "Maps in-force status, consolidation, amendment and repeal relations and authority attribution onto the local status and supersession model where the decision is a normative instrument. The ELI identifier, when assigned, ranks as a governed global identifier and does not displace the master-system identifier.", "required": false, "source_refs": [ "SRC-031" ] }, { "target": "W3C Verifiable Credentials Data Model v2.0", "relation": "ALIGN", "purpose": "Maps issuer attribution, validity windows and revocation or suspension status onto the local authority and status model where a decision is projected as a verifiable credential. Credential formats and proof mechanisms remain a projection concern.", "required": false, "source_refs": [ "SRC-026" ] }, { "target": "OASIS Akoma Ntoso v1.0 Part 1 XML Vocabulary", "relation": "ALIGN", "purpose": "Maps FRBR-style work, expression, manifestation and item identification, lifecycle events, temporal data and top-level person, role and organisation references onto local version identity and participation structures for legislative and deliberative contexts.", "required": false, "source_refs": [ "SRC-012" ] } ], "serviceLayers": { "dimension": { "owner_package_requirements": [ "The adopting Dimension must name one accountable decision-governance owner package and its owning organisational unit before any governance record is created; unowned governance records are invalid.", "The owner package must publish the decision-rights or mandate catalogue it operates, the status vocabulary and permitted transitions, and the retention schedule authority under which it disposes of decision records.", "The owner package must declare the jurisdictional profiles in force — at minimum the access-to-documents regime, the data-protection regime and any sector records legislation — and the local overrides each imposes on the default rules in this model.", "The owner package must nominate a custodian for every stored projection and a reviewer or review forum for every decision class subject to periodic review, and must keep those nominations current through the role handover function." ], "namespace_guidance": "Use one stable, dereferenceable namespace owned by the adopting Dimension, shaped as {dimension-base-iri}/decision-governance/{local-id}, where {local-id} is the master identifier or the Dimension-assigned ULID. Never encode a date, a status, an access class or an organisational unit into the namespace path, because all four change while identity must not. Where a governed global identifier such as an ELI-style legal identifier exists it is recorded as an alias and may be minted in its own namespace, but it never replaces the master identifier in the local namespace. Artefact-kind discriminators (GOVLED, GOVACC, GOVDIS) are part of the artefact name, not of the namespace.", "registry_links": [ "Vercy registry entry vr.wm-knw-010 for model identity, entry kind, status and review state", "planning/VERCY-MODEL-RELATIONS.csv for the authoritative relation ledger governing composition and ownership boundaries", "The adopting Dimension's decision authority and delegation register, resolved through dr-gov-art-authority-register", "The adopting Dimension's retention schedule register and its issuing records authority" ] }, "canon_and_patch": { "canonicalization_rules": [ "Serialize with deterministic member ordering and a declared character encoding; the digest in the integrity rule is computed over this canonical form and the canonical form must be named alongside the digest.", "Normalise every timestamp to RFC 3339 with explicit seconds and an explicit offset or Z, and retain the originally submitted offset as a separate field so that local civil time is not lost by normalisation.", "Represent identifiers verbatim as issued by their master system, including case; do not case-fold, trim or re-pad a master identifier, and record any locally applied alias separately.", "Store codes as code plus code-system reference; never store a display label as the value. Absent and empty are distinct: an absent field means not stated, an empty collection means stated as none." ], "patch_rules": [ "The lifecycle ledger is append-only. Entries are never edited or removed; a correction is a new entry that references the entry it corrects and states the reason.", "An approved or effective version is immutable. Substantive change is expressed by issuing a new version with a supersession reference, never by mutating the prior version; editorial correction to an approved version requires an erratum entry citing the authorising party.", "Status changes only through the governed transition function, which validates the transition against the declared transition table and records acting authority, ground and effective instant.", "Every patch carries the acting agent, the reason code, and both the event instant and the ingestion instant; a patch missing any of these is rejected.", "Access classification and retention bindings are versioned, not overwritten: a reissued access schedule or disposition record increments its own zero-padded sequence and retains its predecessor." ], "compatibility_rules": [ "Adding an optional data element, an optional question or a new alignment is a minor, backward-compatible change.", "Removing or narrowing a status value, an access class, a role or a code system, or making an optional element required, is breaking and requires a new major model version and a migration note.", "Changing an alignment mapping is non-breaking where local semantics are unchanged; the alignment record must state the source version it maps to so that a target standard's revision does not silently alter local meaning.", "Versioning applies to the model and to individual decision records independently: a model version change never rewrites the version identifiers of existing decision records.", "Consumers must ignore unknown members rather than fail, but must not infer a default for an absent required element." ] }, "artifact_rules": { "identity_priority": [ "Authoritative master-system identifier: the identifier issued by the system of record for the decision — typically the deciding body's decision or minute register number — used verbatim as the decision master identifier and never re-minted locally.", "Governed global identifier or IRI assigned by a recognised registry, such as an ELI-style legal identifier or another registry-governed IRI, used when no master-system identifier exists or recorded as an alias when one does.", "UUID or ULID minted by the adopting Dimension, flagged as Dimension-assigned, used only when neither of the above exists; if a master identifier is later issued it is adopted as primary while the Dimension-assigned identifier is retained permanently as an alias.", "A date, a meeting name, a title, an agenda item number, a file path or a bare sequence is never an identifier and must not be used as one, alone or in combination." ], "timestamp_rule": "All timestamps are RFC 3339 date-time values that include explicit seconds and an explicit UTC offset or the Z designator; a bare date, a date-time without seconds, and a date-time with no offset are all rejected. The offset -00:00 is used only where the instant in UTC is known but the local offset genuinely is not. Event time (when the governance act occurred, such as the decision being taken or a hold being imposed), observation time (when the fact was witnessed, attested or reported by a named party) and ingestion time (when the Vercy record captured it) are stored in separate fields and are never merged, back-filled from each other, or defaulted to the current clock; where a fact has only one meaningful time axis, the unused axes are left absent rather than duplicated.", "serial_naming_rule": "Every serial governance artefact is named {decision-master-identifier}-{artefact-kind}-{NNNN}: the decision master identifier resolved by the identity priority above, a fixed artefact-kind discriminator (GOVLED for the lifecycle and change ledger, GOVACC for the access and disclosure schedule, GOVDIS for the review, hold and disposition record), and NNNN, a zero-padded monotonic sequence whose scope is explicitly the decision master record. Sequences start at 0001, are never reused after supersession or deletion, and are never derived from a date, timestamp, retention year, review cycle or period label. A gap in a sequence must be explained by a lifecycle ledger entry. Non-serial artefacts, such as the authority and delegation register, take their identity from the identity priority instead and are keyed by authority plus mandate instrument.", "integrity_rule": "Each governed version and each serial artefact version records a named digest algorithm, the digest value computed over the declared canonical form, the sealing agent and the sealing instant, plus any signature, seal or timestamp-token reference evidencing non-repudiation. Verification compares a retrieved projection against the stored digest and writes a verification outcome entry; a failure quarantines the projection and raises a notification, and the sealed version is never corrected in place. Vercy stores integrity values and references only: generating, protecting, reviewing and retaining audit trails, and enforcing any control derived from them, belong to the referenced records and audit systems." }, "policies": [ "Authority before content: no decision record may reach an approved or effective status without a resolvable deciding authority and mandate binding whose validity window covers the decision event time.", "Reference, do not reproduce: approval activities, carrying records, legal judgments, claims, evidence, policies and executable rules are held as identifiers with a binding role and decision-specific parameters only; their internal structures, lifecycles and operational functions are never mirrored here.", "No local evaluation or enforcement: this model records assertions, classifications and instructions. Threshold evaluation, access enforcement, rule execution, obligation discharge and audit-trail generation are performed by external systems, and referencing such a system never transfers its semantics to this model.", "Immutability after approval: approved and effective versions are immutable; change is expressed through new versions, supersession and append-only ledger entries, so that what was relied upon at a past instant remains reconstructable.", "Tombstone-first disposition: content is superseded or tombstoned rather than erased unless an authorised erasure rule applies, and identity, authority, status and disposition facts survive every disposition that is not itself an authorised erasure of those facts.", "Declared profiles, no global claims: jurisdiction-specific access, integrity and retention duties are recorded as named profiles or alignments; the model never asserts conformance to a standard or a legal regime without evidence recorded in the adopting Dimension." ], "crud": { "read": [ "Read is governed by the access classification held for each addressed part; the outcome, the rationale reference set and the governance metadata may carry different classes and are evaluated separately.", "Register-level metadata declared in the discoverable metadata set — master identifier, subject reference, deciding authority and decision date — remains discoverable to authenticated readers even where content is withheld, unless a sensitive-item rule restricts the existence of the item itself.", "Where an exemption covers only part of a record, readers receive the parts not covered, produced through the declared redaction profile; a partial release is recorded as a disclosure event in the lifecycle ledger.", "Reads resolve references without dereferencing target-model content: a reader receives approval, evidence, record and judgment identifiers, not copies of those records.", "Point-in-time reads must be answerable: given an instant, the reader can determine which version was authoritative and which status applied." ], "create": [ "Only a party holding the governance owner or decision authority role may create a governance record, and creation must bind the decision master identifier, the deciding authority and the mandate reference.", "Creation records the ingestion instant separately from the decision event instant, and a draft record must not carry an effective status or an effective instant.", "A newly created record with no master-system identifier is minted a Dimension-assigned ULID that is flagged as such and retained as an alias if a master identifier later appears.", "Creation opens the lifecycle ledger for the decision with an initial entry naming the creating agent, the capture mode and the source system where the record was imported or migrated." ], "update": [ "Governance metadata on a draft may be updated in place; once a version is approved or effective it is immutable and change proceeds by new version and supersession reference.", "Every update writes an append-only lifecycle ledger entry with acting agent, reason code, significance code and the separated event and ingestion instants; silent edits are rejected.", "Status changes only through the governed transition function and only where the transition is permitted from the current status and the acting authority is competent for it.", "Corrections to approved content are recorded as errata that cite the authorising party and leave the corrected version retrievable in its original sealed form.", "No update may weaken an access classification or shorten a retention binding without a recorded authorisation from the classifying or records authority." ], "delete": [ "Deletion is tombstone-first and supersession-first: superseding, revoking or expiring a decision never removes it, and content removal requires a disposition authorisation recorded in dr-gov-art-disposition-record.", "Disposition is blocked while any legal hold or disposition freeze is in force, and while a review, appeal or reconsideration is open against the decision.", "When an authorised disposition removes content, the surviving tombstone retains the decision master identifier and any aliases, the version identifier, the deciding authority reference, the status at disposition, the disposition action code, the disposition authorisation reference and the disposition instant, together with the lifecycle ledger entries that evidence the disposition.", "Execution of destruction, transfer or migration is not performed by this model: it is owned by the carrying record model WM-REC-010 and the adopting Dimension's records authority operating under its published retention schedule, and this model records only the authorisation, the request and the evidence returned.", "An authorised erasure rule — for example a data-protection erasure duty, subject to its own statutory exceptions for legal obligations, public-interest tasks, archiving and the defence of legal claims — may require removal of identifying content from the tombstone itself; where it applies, the erasure authorisation reference, its legal basis and the resulting reduced tombstone are recorded, and the surviving audit and integrity facts held by the external audit system are not deleted by this model.", "Deletion of an entire lifecycle ledger, access schedule or disposition record is never permitted while the decision record exists; these artefacts are disposed of only as part of the decision's own authorised disposition." ] }, "roles": [ { "name": "Decision authority", "responsibilities": [ "Take the decision within the mandate and record the authority and mandate binding", "Attest satisfaction of the applicable decision rule and participation threshold", "Authorise status transitions such as suspension, revocation and reopening", "Determine whether a declared interest is disqualifying where the rules assign that determination to the chair or presiding office" ] }, { "name": "Governance owner", "responsibilities": [ "Own the decision record as a governed asset and account for its completeness", "Maintain the authority and delegation register entries relied on by the decision", "Approve role handovers and nominate custodians and reviewers", "Declare the jurisdictional profiles applied to the record" ] }, { "name": "Records steward", "responsibilities": [ "Maintain the accuracy of governance metadata, including status, effective dating and version chain", "Write and quality-check lifecycle ledger entries and errata", "Bind the record to the correct retention schedule item and trigger", "Prepare disposition requests and record the evidence returned" ] }, { "name": "Custodian", "responsibilities": [ "Hold and protect a stored projection of the record", "Compute and store digests over the declared canonical form and report verification outcomes", "Execute or relay disposition instructions to the executing records authority", "Preserve tombstones and integrity references after content removal" ] }, { "name": "Reviewer or appeal forum", "responsibilities": [ "Conduct scheduled and triggered reviews by the recorded due instant", "Receive appeals and reconsideration requests within the applicable window", "Record the review outcome and its relation to the original decision without overwriting it", "Refer matters to a judicial forum where the escalation tier requires it" ] }, { "name": "Access and disclosure authority", "responsibilities": [ "Set and revise access classifications and exemption grounds", "Approve partial access and the redaction profile applied", "Release the decision proactively when an embargo lapses", "Apply personal-data minimisation before any disclosure" ] } ], "access": { "default_rule": "Deny by default at finding and artefact scope. A reader obtains content only where an access class explicitly grants it for the addressed part; register-level metadata in the declared discoverable set is available to authenticated readers unless a sensitive-item rule restricts even the existence of the record. Enforcement is performed by the adopting Dimension's access-control system; this model supplies the classification and the disclosure instruction only.", "scopes": [ "bundle", "layer", "finding", "artifact" ], "exceptions": [ "Proactive publication: once an embargo lapses and no hold or open appeal blocks release, the decision is published through the recorded channel by the named release authority.", "Statutory access request: a request under an access-to-documents regime is answered by releasing the parts not covered by an exemption, applying the redaction profile, and giving written reasons for any refusal together with the available remedies.", "Sensitive-item handling: items classified as sensitive are handled only by authorised personnel and are released only with the originating authority's consent, and their register references are worded so as not to disclose the protected content.", "Legal hold and discovery access: a party conducting a hold or discovery obtains access for that purpose under a recorded authorisation, without altering the record's classification.", "Emergency break-glass: an urgent operational need may be met by a time-boxed override that is recorded with the requester, the justification and the expiry instant, and that requires post-hoc ratification by the access and disclosure authority." ], "audit_requirements": [ "Every access decision records the requester, the addressed scope, whether access was granted, refused or partial, and the exemption ground where refused or redacted.", "Every break-glass override records its justification, its expiry instant and the outcome of its post-hoc ratification.", "Disclosure events, embargo lapses and classification changes are written to the lifecycle ledger with separated event and ingestion instants.", "Audit records themselves are generated, protected and retained by the external audit system identified in the audit trail reference; this model retains only the correlation identifier and a coverage note stating which governance acts are expected to appear there." ] }, "agents_bootstrap": { "filename": "AGENTS.md", "required_fields": [ "Name", "Type", "Specification URL", "Storage type URL", "Interface URL", "Processes URL", "Model ID", "Registry ID", "Governance owner", "Default access class", "Identity priority summary", "Timestamp rule summary", "Relation ledger URL" ], "read_order": [ "AGENTS.md first, to obtain Name, Type and the four URLs before touching any record", "Specification URL, to load scope, boundaries, status vocabulary and permitted transitions", "Relation ledger URL, to confirm which concepts are owned here and which belong to referenced models", "Storage type URL, to learn the projection in use and its canonical form, whether that is a document store, a graph, a repository or a collection", "Interface URL, to learn how identifiers resolve and how reads are scoped by access class", "Processes URL, to learn the governed functions, their preconditions and the roles competent to invoke them", "Only then read or write records, resolving the decision master identifier before any artefact name is constructed" ] } }, "coverage": { "claim": "Audited as a bounded reviewable single-provider draft: 33 sources, 9 bundles, 18 layers, 39 findings, 175 questions across 27 question kinds, 14 artefacts and 21 functions, routed across decision core, rationale and governance with one designated owner per checklist dimension. Coverage is claimed only against that routed checklist, the two registered ledger relations and the declared jurisdictional profiles; it is not universal. Three dimensions are self-declared gaps (causal and counterfactual rationale, reason weighting and aggregation semantics, participation and conflict-of-interest structure), one is not-applicable by boundary, and twenty known omissions plus sixteen regional assumptions remain open. No independent second-provider corroboration exists, and no source URL or version pin was re-resolved in this no-tools audit. No contradiction found in this audit prevents publication of a reviewable draft carrying the holds below.", "confidence": "medium", "checklist": [ { "dimension": "identity", "status": "covered", "notes": "Master-system identifier first, then a governed global identifier, then a Dimension-minted UUID or ULID; version discriminator, supersession links and labelled non-authoritative alternate identifiers. DMN's tDMNElement identifier is an xsd:ID scoped to one interchange document and is treated as an alignment only, never as a source of stable identity." }, { "dimension": "relationships", "status": "covered", "notes": "Typed references for the triggering issue, sub-decisions, upstream decisions, required inputs, knowledge sources, affected subjects, addressed and raised concerns, superseded decisions and follow-on decisions. Each outbound link is a reference or alignment, never an embedded copy." }, { "dimension": "alternatives and baseline", "status": "covered", "notes": "Append-only registration of alternatives with kind codes, variant links and a profile-conditional no-action or business-as-usual baseline, grounded in the Green Book 2026 mandatory benchmark and the 2024 CFR edition of 40 CFR 1502.14(c)." }, { "dimension": "criteria and weighting", "status": "covered", "notes": "Criteria with definition, measure, scale, unit and preference direction; mandatory versus compensatory classification; optional weights with a declared method. Named MCDA methods and their axioms remain optional external bindings." }, { "dimension": "evaluation and comparison", "status": "covered", "notes": "Per-criterion results, aggregation, ranking and sensitivity recorded as serially numbered append-only rounds, with a comparison-matrix artefact. Runtime execution of any decision service that produced the numbers is explicitly excluded." }, { "dimension": "outcome and disposition", "status": "covered", "notes": "Selected alternative, disposition code, immutable outcome statement, balanced factors, divergence explanation and separately identified preferable alternative, following the record-of-decision content pattern without owning the published record." }, { "dimension": "measurement", "status": "covered", "notes": "Constraint limits, criterion measures, scales, units, preference directions, scores, aggregates and monetisation basis are explicit; quantities carry units and monetised values carry a currency and price base." }, { "dimension": "jurisdictional profile", "status": "covered", "notes": "Decision type binds a profile that declares required content elements and validation rules. No global default profile is asserted, and profiles are modelled as an opt-in MIX-IN rather than as built-in requirements." }, { "dimension": "uncertainty and confidence", "status": "covered", "notes": "Confidence qualifiers carry a named scale, an explicit subject (evidence, inference or outcome) and stated knowledge limits, and are kept inline with what they qualify so a figure cannot circulate without its scale." }, { "dimension": "contestation and dissent", "status": "covered", "notes": "Counter-arguments, counter-evidence, objection attribution, appended dispositions, argument-level defeated markings and separately attributed minority opinions are modelled and protected from overwrite, with adjudication authority left to governance." }, { "dimension": "explanation faithfulness", "status": "covered", "notes": "Every rendition carries the rendered rationale revision, derivation method, omission record and a non-suppressible generated-content label, matching the requirement that an explanation reflect the process that produced the output rather than a plausible reconstruction." }, { "dimension": "causal and counterfactual reasoning", "status": "gap", "notes": "Modelled only as assertion records with a strength qualifier, cited basis and competing-explanation note. No consulted primary source defines a normative structure for counterfactual rationale, so this finding is marked as a gap rather than presented as canonical, and the model never upgrades an association to an asserted cause." }, { "dimension": "reason weighting and aggregation", "status": "gap", "notes": "Asserted reason weights and decision-local evidential weights are carried with a mandatory scale reference, but no consulted primary source supplies a governed weighting or aggregation semantics. Weights are therefore recorded as asserted values only and must not be summed, compared across decisions or treated as calibrated." }, { "dimension": "lifecycle", "status": "covered", "notes": "A controlled status vocabulary covers draft, submitted, approved, effective, conditional, suspended, superseded, revoked, expired and reopened with a transition table, competent authorities and separated approval, publication and effective instants. RFC 2026 supplies staged maturity, retirement and appeal precedent; ELI supplies in-force, consolidation and repeal relations." }, { "dimension": "temporal", "status": "covered", "notes": "All instants are RFC 3339 with seconds and an explicit offset or Z; event, observation and ingestion times are separate fields; validity windows, embargo lapse, appeal windows and retention triggers are modelled explicitly. Retrospective effect is flagged where the effective instant precedes approval." }, { "dimension": "provenance", "status": "covered", "notes": "Attribution, derivation, attestation and capture mode align to PROV-O qualified patterns; the append-only ledger records acting agent, reason and authority for every governance act. Cited claims and evidence are identifier-only references." }, { "dimension": "ownership", "status": "covered", "notes": "Owner, steward, custodian and reviewer are distinguished, with handover events and residual accountability, and separated from the deciding authority that owns the substance. Six service-layer roles carry explicit responsibilities." }, { "dimension": "validation", "status": "covered", "notes": "Quorum and decision-rule satisfaction is recorded as an attributed assertion by the presiding authority rather than computed here; integrity verification produces an outcome entry without mutating the sealed version; approval defects have codes and remediation paths." }, { "dimension": "access", "status": "covered", "notes": "Deny by default at all four scopes, per-part classification, exemption grounds with provision references, partial access by redaction, discoverable register metadata and five named exceptions including a ratified break-glass override. Enforcement is explicitly external." }, { "dimension": "retention and deletion", "status": "covered", "notes": "Retention schedule binding with an explicit start trigger, legal hold blocking disposition, tombstone-first deletion with an enumerated surviving field set, named executing owner (WM-REC-010 and the Dimension records authority), archival transfer for permanent-value records, and an authorised-erasure path with its statutory exceptions." }, { "dimension": "interoperability", "status": "covered", "notes": "Six alignments recorded as alignments only — ISO 15489-1, PROV-O, ELI 1.5, VC Data Model 2.0, Akoma Ntoso and NARA's requirement categories — each stating that local semantics prevail on divergence and that no conformance is claimed." }, { "dimension": "authority and mandate", "status": "covered", "notes": "Mandate instrument, scope limits, validity window, succession and ultra vires handling are modelled, with the authority register as the resolution point. Grounded in ISO 15489-1 assigned responsibilities, ELI responsibility_of_agent and passed_by, and the Council's Rules of Procedure." }, { "dimension": "participation and conflict of interest", "status": "gap", "notes": "Participation mode, proxy representation and quorum attestation are grounded in Council Decision 2009/937/EU, which is jurisdiction-specific. No cross-domain primary standard was found for a conflict-of-interest register schema or for recusal effects on thresholds; the OECD conflict-of-interest guidance could not be retrieved during this pass, so this structure is marked as a gap rather than presented as canonical." }, { "dimension": "integrity and audit linkage", "status": "covered", "notes": "Digest algorithm and value over a declared canonical form, sealing agent, signature or timestamp-token reference, quarantine-on-failure and an external audit correlation identifier. Audit generation, protection and retention are stated as externally owned in line with the NIST control families." }, { "dimension": "implementation and outcome monitoring", "status": "not-applicable", "notes": "Held only as identifier-only follow-up obligation references with the owning model named. Implementation progress, delivery data and outcome measurement are outside this model's boundary by design and no local structure is offered for them." } ], "known_omissions": [ "Group aggregation and voting rules, including tie-breaking and dissent recording, which the sources consulted do not settle.", "Probabilistic decision structures such as decision trees, influence diagrams and expected-utility calculations, treated as optional method bindings.", "Named multi-criteria methods and their axioms, including AHP, ELECTRE, PROMETHEE and TOPSIS.", "Domain rule content such as clinical protocols, credit scorecards, procurement award formulas and sentencing guidelines.", "Appraisal cost conventions such as discount rate, price base and currency, which are jurisdiction-specific and change independently of this model.", "Machine-learning model cards or feature attributions for automated decisions, which belong to a model-governance sibling rather than to decision content.", "No normative structure for counterfactual or causal rationale was found in any consulted primary source; the corresponding finding is flagged as a gap and carries assertions only.", "Deliberation and dialogue modelling — turn taking, protocol state, quorum and voting mechanics — is omitted; AIF dialogue extensions and inference-anchoring approaches were not adopted.", "Formal defeasible-reasoning semantics (abstract argumentation frameworks, structured argumentation calculi) are not adopted; conflict and defeat are recorded as asserted annotations with no acceptability computation.", "Domain-specific reason taxonomies — legal grounds of decision, clinical reasoning categories, procurement justification codes — are left to the adopting Dimension's published vocabulary rather than enumerated here.", "Multilingual equivalence between language versions of a stated reason is not modelled beyond retaining the original language tag and forbidding silent translation.", "Normative clauses of ISO/IEC/IEEE 42010-2022 and ISO/IEC/IEEE 15026-2 on rationale and assurance cases are paywalled and were not consulted; only publicly available scope text was used, so architecture-description rationale requirements are not reflected in detail.", "Quantitative confidence calibration methods and the mapping between verbal qualifiers and numeric ranges are required to be published but no default mapping is supplied.", "No controlled vocabulary is fixed for decision rules, recusal actions or reopening grounds; the adopting Dimension must publish one, and the codes here are placeholders bound to a code-system reference.", "ISO 16175-1:2020 and the ISO 23081 metadata-for-records series were not directly consulted because their texts are paywalled; the records-metadata entity model, including an explicit mandate entity, is therefore inferred from ISO 15489-1 and NARA rather than from the most specific standard available.", "The OECD guidance on managing conflict of interest in the public service could not be retrieved during this pass, so conflict-of-interest structure rests on a single EU procedural instrument and a data-protection instrument.", "No structure is offered for meeting minutes or proceedings as such; where minutes exist they are referenced, and the model does not attempt to be a substitute account of a sitting.", "Voting tallies are not modelled beyond an attested counted-and-required pair, because the dissent and objection content is owned by the reasoning area and a full ballot structure would duplicate it.", "Cost, effort and resourcing consequences of a decision are not modelled at all; they belong to planning and financial models outside this boundary.", "Cryptographic scheme profiles, key management and trust anchors are referenced but not specified; they belong to the adopting Dimension's security model." ], "conflicts": [ "OMG DMN models a reusable decision definition inside a decision requirements graph, whereas WM-KNW-010 governs an instance-level decision taken on a specific occasion. DMN carries no version attribute on Decision and its identifier is an xsd:ID scoped to one interchange document, so DMN identifiers cannot satisfy the Vercy identity priority. Recorded as an alignment conflict, not resolved by adopting DMN identity.", "DMN's Decision element has no representation for alternatives, criteria, weights, rationale or disposition; it carries question, allowedAnswers, requirement links and an expression. DMN therefore supports the framing findings but cannot support the alternatives, evaluation or outcome findings, which is why those are grounded in appraisal and decision-analysis sources instead.", "The no action alternative requirement in 40 CFR 1502.14(c) and the record-of-decision content in 40 CFR 1505.2 are cited from the 2024 CFR annual edition. CEQ removed 40 CFR parts 1500 to 1508 by interim final rule at 90 FR 10610, published 25 February 2025 and effective 11 April 2025, directing agencies to rely on their own procedures. These are therefore cited as a historical normative content pattern, not as currently binding government-wide requirements.", "HM Treasury's supplementary guidance restricts multi-criteria decision analysis to longlist appraisal and warns that it is resource-intensive and needs specialist practitioners, which conflicts with widespread practice of weighted scoring at shortlist stage. Weights are consequently modelled as optional and never as a completeness requirement.", "GDPR Article 22 uses decision in a legal-effects sense that coincides with neither DMN's technical decision node nor ISO/IEC/IEEE 42010's architecture decision. The model does not reconcile these senses; it makes decision type and jurisdictional profile explicit so the intended sense is always recoverable.", "ISO/IEC/IEEE 42010 treats architecture rationale as a first-class element distinct from the decision, while appraisal guidance folds justification into the appraisal record itself. This model follows 42010 and keeps rationale separate, which means a Green Book style appraisal summary maps across two areas of WM-KNW-010 rather than one.", "Structured-argument metamodels and the Argument Interchange Format both make inference and conflict first-class but with different primitives — asserted inference and evidence links with counter flags versus rule-application, conflict-application and preference-application nodes — and no normative crosswalk between them exists. This model records element roles abstractly and asserts only role-level alignment.", "Goal Structuring Notation's Justification, the assurance metamodel's ArgumentReasoning and the Argument Interchange Format's rule-application node are frequently treated as equivalent in practice but are not defined equivalently; conflating them would silently change what is being asserted.", "Citation typing is rhetorical and bibliographic while certainty-of-evidence grading is methodological. A citation type such as 'cites as evidence' carries no quality claim, yet tooling routinely presents the two together; this model forbids reading a citation type as a grade.", "Web Annotation motivation values do not include an evidential-citation motivation, so citation type must come from a separate citation vocabulary while locators come from the annotation model; the two vocabularies are combined without a published joint profile.", "Explainability guidance requires explanations to be accurate to the generating process, while regulation frames the explanation as an entitlement of an affected person expressed in accessible terms. These pull in different directions for technical decisions, and the model resolves it by requiring an explicit omission record rather than by preferring one over the other.", "Records doctrine treats a captured record as fixed, while ELI and Akoma Ntoso treat a legal text as an expression that is consolidated and versioned over time. Resolved here by making versions immutable once approved and expressing all change as new versions with supersession links, so both readings hold, but adopters projecting to a records system that forbids multiple live expressions will need a profile.", "A data-protection erasure duty can require removal of identifying content that a retention schedule and a legal hold require to be kept. The model's ordering — hold blocks disposition, and authorised erasure may reduce a tombstone — reflects the statutory exceptions for legal obligations, archiving and defence of legal claims, but the resolution is jurisdictional and must be decided by the adopting Dimension's records and privacy authorities, not by this model.", "Verifiable-credential revocation semantics imply a status that can be flipped by the issuer at any time, whereas records doctrine requires the superseded state to remain retrievable. Resolved by treating credential status as a projection of the governed status, never as its master.", "Access-to-documents regimes create a presumption of disclosure with enumerated exemptions, while security control catalogues start from deny-by-default. The model defaults to deny and treats proactive publication and statutory requests as recorded exceptions, which is conservative but will understate disclosure duties for public bodies unless the Dimension declares the access profile.", "NARA's requirement set separates mandatory from preferred requirements for a records system, while this model states governance obligations for a decision record; an adopter should not read a 'Must Have' system requirement as satisfied merely because the corresponding governance field exists here." ], "regional_assumptions": [ "The 40 CFR material is United States federal environmental review practice only, and is additionally superseded as described in the conflicts.", "The Green Book and its multi-criteria supplement are HM Treasury guidance for United Kingdom central government and carry no authority elsewhere.", "GDPR obligations on automated decision-making apply to processing within the EU and EEA scope; equivalent regimes in other jurisdictions are not assumed to impose the same content requirements.", "NASA's decision analysis material is United States federal aerospace systems-engineering practice, used here as a method pattern rather than as a mandate.", "No global default jurisdictional profile is asserted. An adopting Dimension must bind a profile per decision type, and an unbound decision type carries no profile-required content elements.", "Language, locale and script of the decision question and outcome statement are recorded per decision; no single working language is assumed.", "The duty to state reasons cited here is a Treaty obligation applying to EU legal acts; it is used as evidence that a stated reason is a governed object, not as a claim that private or non-EU decisions carry the same duty.", "The right to explanation cited here applies to deployers of specified high-risk AI systems under EU law, with statutory exceptions; it does not establish a general right to explanation for all decisions in all jurisdictions.", "OMB Circular A-4 is United States federal guidance for regulatory analysis, archived in the 2003 form consulted, and has been revised and contested since; it is cited as an illustrative public-authority requirement for sensitivity analysis and assumption transparency, not as a globally binding rule.", "The certainty-of-evidence framework originates in health-evidence appraisal and is rendered here through a United States public-health handbook; transfer to engineering, legal or commercial decisions is by analogy and must be declared by the adopting Dimension.", "Deliberative bodies differ in whether dissent is recorded, attributed or published at all; this model requires that dissent be preserved where it exists but does not assume any jurisdiction's disclosure rules.", "Regulation (EC) No 1049/2001, Council Decision 2009/937/EU and Regulation (EU) 2016/679 are EU instruments used as profiles. They illustrate exemption grounds, partial access, quorum, proxy voting, written and silence procedures, storage limitation and erasure exceptions; they are not asserted to apply outside the EU.", "NARA's Universal ERM Requirements and NIST SP 800-53 are United States federal profiles. Their requirement categories and control families inform structure but confer no obligation on non-US adopters.", "ELI-style identifiers and Akoma Ntoso structures presume a legislative or deliberative context; for corporate or operational decisions they are optional alignments only.", "No sources from Asia-Pacific, African, Latin American or Middle Eastern records or governance regimes were consulted in this pass, so region-specific duties there are unassessed.", "The tiered appeal model drawn from RFC 2026 reflects one standards organisation's practice; statutory appeal rights, standing rules and time limits vary widely and must be supplied by the adopting Dimension." ], "adversarial_checks": [ "Searched for a counterexample where a decision has no genuine alternatives, such as a mandatory statutory determination. The model accordingly permits an alternative set of size one plus an optional baseline, makes weights and aggregation optional, and never requires a ranking to exist for an outcome to be complete.", "Tested whether DMN alone could carry the whole model. It cannot: the Decision element lacks alternatives, criteria, weights, rationale and disposition, and its identifier is document-scoped. DMN was therefore demoted to an alignment with a recorded conflict rather than adopted as a conformance target or an identity source.", "Checked every finding and function against the two ledger relations and the six neighbouring models. Evidence appears only as a reference with a citation role, participants and approval appear only as an outbound reference, the carrying record appears only as a reference, and no function performs evaluation execution, enforcement or audit logging. The decision-rule finding was rewritten to record the rules relied upon plus a versioned source reference, with execution explicitly assigned to a decision service.", "Tested whether the no-action requirement is universal. It is normative in the Green Book as the business-as-usual benchmark and was normative in the 2024 CFR edition of 40 CFR 1502.14, but is absent from DMN and from 42010. It is therefore modelled as a profile-conditional requirement with a recorded justification for absence, not as a globally mandatory field.", "Checked the machine-gate conditions directly: identity priority names the authoritative master-system identifier first; the timestamp rule contains RFC 3339, seconds and offset and separates event from observation or ingestion time; the delete rules state tombstone, retention and erasure ownership; all local identifiers begin with dr-core-, are unique and contain no date-like component; every finding either declares artefacts with a null rationale or declares an empty artefact array with a substantive rationale, never both; question texts are distinct across the model and span eighteen distinct question kinds.", "Probed whether the outcome finding smuggles in approval semantics. The disposition code names what the decision content states, and the function effects place approval execution, publication, filing and enforcement with the referenced models, so a consuming agent is instructed not to infer approval or legal effect from this model alone.", "Checked every finding against the WM-KNW-008 reference rationale: no finding reproduces evidence identity, acquisition, custody, grading methodology or retention. The one finding that assesses evidence is explicitly scoped to a decision-local appraisal, carries the referenced certainty grade unchanged with a read time, and is inline-only so it cannot circulate as an evidence assessment report.", "Tested whether the incoming WM-ACT-032 relation had leaked ownership: no bundle, layer, finding or function mentions change-request intake, triage, workflow or closure, and the composition entry states that this model only exposes resolvable identifiers.", "Tested whether 'explainable decision trace' had become an audit trail. The trace finding, the rendition artefact and the access audit requirements all state that audit trails, access logging and disclosure approval belong to the `governance` split; the render function's effects say it is not an audit record and discharges no regulatory duty.", "Tested whether the assumption and sensitivity function had become a runtime evaluator or enforcement engine. Its effects restrict it to annotating status and listing stale references, explicitly excluding executing analysis, re-running models, enforcing outcomes and changing decision state, with escalation left to governance.", "Searched for the counterexample that argument standards imply an evaluation semantics. Assurance-case and argument-interchange sources describe auditable claims, arguments and evidence and asserted inference, not entailment computation, so the model records asserted inference and conflict and exposes no validity, soundness or acceptability function.", "Checked that no structure asserts truth, legality or optimality: the policy layer forbids it, the render function's effects repeat it, and the causal finding is constrained to assertions with a strength qualifier and competing-explanation note rather than causal conclusions.", "Rejected an attractive but unsupported structure: a 'rationale quality score' aggregating reason weights and evidence grades. No consulted primary source supports aggregation across those scales, so weighting is recorded as asserted values only and the aggregation question is logged as a gap.", "Ownership sweep against the relation ledger: every bundle, layer, finding and function was re-read against the WM-ACT-032 and WM-KNW-008 rationales and against each outgoing reference. Approval execution, record disposal execution, evidence appraisal, claim structure, judicial reasoning, change-request lifecycle and rule evaluation were each found to be referenced only; where a draft concept drifted towards ownership — notably follow-up obligations and audit trails — it was reduced to an identifier-only reference and restated in out_of_scope or a boundary note.", "Evaluation and enforcement test: the quorum function was deliberately renamed to attest rather than verify, and its effects state that no threshold computation or enforcement is performed. The integrity function verifies a digest but writes an outcome entry rather than acting on it, and the access layer emits classifications and instructions while enforcement is assigned to an external system.", "Artefact collision test: the four artefacts introduced here (dr-gov-art-authority-register, dr-gov-art-lifecycle-ledger, dr-gov-art-access-schedule, dr-gov-art-disposition-record) were checked against the ten pre-existing core and reasoning artefact identifiers; none is re-created, renamed or shadowed, and dr-rsn-art-dissent-record is referenced by identifier only.", "Identity falsifiability test: every artefact identity strategy and every local identifier was checked for date-derived or period-derived components. Serial names use a zero-padded sequence scoped to the decision master record, and the timestamp rule explicitly refuses bare dates, second-less times and offset-less times.", "Exclusive representation test: each of the thirteen findings was checked to have either non-empty artefacts with a null rationale or an empty artefacts array with a substantive rationale; nine are inline-only with reasons that turn on reference-graph shape, personal-data minimisation or avoidance of duplicating an externally owned document.", "Counterexample search for structure: a separate voting-record structure, a minutes artefact, an interests-register artefact and an implementation-tracking layer were each considered and rejected, because in every case an external owner already holds the concept and no primary source supported placing it here." ] }, "researchAdjudication": { "providerMode": "single-provider-waiver", "activeProviders": [ "claude" ], "waivedProviders": [ "grok" ], "providerPolicy": { "contract_version": "1.0.0", "mode": "single-provider-waiver", "effective_at": "2026-08-29T09:06:27Z", "scope": "Queued subject-model research from WM-XCT-013 onward", "active_providers": [ "claude" ], "waived_providers": [ { "provider": "grok", "authorized_by": "repository owner", "authorized_at": "2026-08-29T09:06:27Z", "reason": "The repository owner explicitly instructed the research queue to continue without Grok after repeated structured-output failures." } ], "review_rule": "Claude-only results require a separate no-tools adversarial audit and remain reviewable drafts with a visible single-provider hold." }, "boundaryDecision": { "entry_kind": "aggregate", "status": "accepted", "rationale": "Two distinct axes must not be conflated. On the record plane the frozen registry value `standalone-mm` classifies how vr.wm-knw-010 sits in the Vercy registry — a self-standing meta-model entry rather than a fragment, mixin or alias of another entry — and it is absent from the subject-model enum, so it is retained as registry metadata and never used as the subject kind. On the subject-model axis `aggregate` is the most defensible schema kind and is affirmed: the model declares one root identity (the authoritative master-system decision identifier), and alternatives, screening records, criteria, evaluation rounds, outcome fields, reasons, argument elements, bindings and ledger entries are addressable only through that root, with artefact names derived from it. Immutability after approval, append-only ledger patching and versioned supersession define a single consistency boundary over that cluster, which is precisely aggregate behaviour. `entity` was rejected as understating an owned multi-part cluster; `event` was rejected because occurrent semantics (convening, deliberation, voting, approval execution) are explicitly assigned to WM-ACT-024 and the subject here persists, versions, is reviewed, reopened, retained and disposed; `relationship`, `classifier`, `mixin`, `registry` and `pattern` each describe referenced machinery (profiles, code lists, external registers) rather than the governed subject." }, "decisions": [ { "concept": "Subject-model entry kind for WM-KNW-010", "disposition": "accepted as aggregate", "rationale": "One root identity governs an internally owned cluster of alternatives, criteria, evaluation rounds, reasons, argument elements and ledger entries under a single immutability and version boundary; entity understates the cluster and event would collide with the occurrent semantics assigned to WM-ACT-024." }, { "concept": "Frozen registry value standalone-mm", "disposition": "retained on the record plane only", "rationale": "It classifies the registry record rather than the modelled subject and is not in the schema enum, so it is preserved as registry metadata while the subject kind is taken independently; the two axes coexist without contradiction and both are stated in the published draft." }, { "concept": "Splitting decision content from rationale into separate models", "disposition": "rejected", "rationale": "Rationale artefacts derive identity from the decision master identifier and rejection reasons hang off core-owned alternatives, so a split would sever the single consistency boundary that justifies the aggregate kind; the existing internal core, reasoning and governance split already routes ownership." }, { "concept": "Merging WM-KNW-010 into registry parent WM-KNW-007 Claim", "disposition": "rejected", "rationale": "A claim asserts that something is the case while a decision selects among registered alternatives and produces a disposition with effect timing, conditions and consequences; generic assertion machinery cannot host alternatives, criteria, evaluation rounds or outcome dating." }, { "concept": "Aggregate root identity and artefact identity priority", "disposition": "accepted", "rationale": "The master-system identifier ranks first, DMN's document-scoped xsd:ID is demoted to an alignment with a recorded conflict, dates, meeting names, titles and file paths are refused as identifiers, and Dimension-assigned ULIDs are retained permanently as aliases when a master identifier later appears." }, { "concept": "Serial artefact-kind discriminator registry", "disposition": "accepted with a pre-promotion condition", "rationale": "The serial naming rule registers only GOVLED, GOVACC and GOVDIS, yet four serial artefacts outside governance — comparison matrix, sensitivity record, dissent record and explanation rendition — rely on unregistered artefact-kind keys, so collision-freedom is asserted rather than demonstrated." }, { "concept": "Sequence scope of dr-rsn-art-dissent-record", "disposition": "deferred for repair before promotion", "rationale": "Its identity strategy scopes the zero-padded sequence per dissenting party while the service-layer rule states the sequence scope is explicitly the decision master record; the two readings produce different names for the same artefact and must be reconciled to one scope." }, { "concept": "Ownership assignments to neighbours absent from the relation ledger", "disposition": "deferred to ledger reconciliation", "rationale": "Delete rules assign disposition execution to WM-REC-010 and boundary notes defer approval, judgment and claim machinery to WM-ACT-024, WM-POL-021 and WM-KNW-007, but the frozen contract registers only the WM-ACT-032 inbound and WM-KNW-008 outbound relations." }, { "concept": "Parent relation to WM-KNW-007 and duplicated identity machinery", "disposition": "deferred", "rationale": "The registry names WM-KNW-007 as parent but the ledger records no such relation, while this model defines its own identity, version discriminator, supersession, provenance and lifecycle; whether that specialises or duplicates parent-owned assertion machinery cannot be settled from the frozen contract alone." }, { "concept": "Missing producing function for the statement of reasons", "disposition": "gap recorded, not repaired in this mode", "rationale": "dr-rsn-art-statement-of-reasons has no governed function that records a reason or issues the statement, whereas every other major artefact has a producing function; add_functions must remain empty under the waiver, so the gap is logged for the next pass rather than filled here." }, { "concept": "Self-referential boundary note naming WM-KNW-010 as its own neighbour", "disposition": "reclassified as an internal split note", "rationale": "Listing the model as its own neighbour pollutes the neighbour contract used for boundary checking, while the content itself is a valid statement about core, reasoning and governance routing and should be labelled as an internal split rather than an external boundary." }, { "concept": "Removed 40 CFR 1502.14 and 1505.2 as content pattern", "disposition": "accepted with a mandatory historical-only label", "rationale": "The pack itself records the CEQ interim final rule removing parts 1500 to 1508 effective 11 April 2025, so the no-action baseline and record-of-decision content bind only as a profile-conditional historical pattern and must never be rendered as a current government-wide requirement." }, { "concept": "Duplicate registration of the GDPR as SRC-005 and SRC-033", "disposition": "accepted with an expression label", "rationale": "The entries are the original OJ text and the 2016-05-04 consolidation of one instrument; without an explicit expression label they inflate apparent source breadth and permit divergent citations of the same article across findings." }, { "concept": "Retention, legal hold and authorised erasure ordering", "disposition": "accepted as a declared jurisdictional profile", "rationale": "Hold blocks disposition, tombstone-first preserves identity, authority, status and disposition facts, and an authorised erasure may reduce the tombstone; the model declines to resolve the statutory conflict globally and assigns it to the adopting Dimension's records and privacy authorities, which is the defensible position." }, { "concept": "Deny-by-default access against access-to-documents presumptions", "disposition": "accepted with a declared-profile condition", "rationale": "The pack concedes the default will understate disclosure duties for public bodies unless an access profile is bound, but discoverable register metadata, partial release by redaction and five named exceptions keep the understatement visible rather than silent." }, { "concept": "No local evaluation, execution or enforcement", "disposition": "accepted after a function-by-function check", "rationale": "The quorum function attests rather than computes, the integrity function writes a verification outcome without acting on it, the assumption re-check only annotates status, and rule execution, access enforcement and audit-trail generation are consistently assigned to external systems." } ], "publicationHolds": [ "Source and version hold: no URL, publication date or version pin was re-resolved in this no-tools audit. All 33 sources must be verified live before publication, with priority on the DMN 1.5 date and document-number mismatch (namespace 2023-03-24 versus formal/24-01-01 versus a stated August 2024 release), the Green Book 2026 edition updated 5 February 2026, CiTO 2.8.2 modified 2026-06-22, and the NIST SP 800-53 Rev. 5 control catalog release 5.2.0 of 27 August 2025.", "Single-provider hold: independent second-provider review is absent by explicit repository-owner waiver recorded at 2026-08-29T09:06:27Z after repeated Grok structured-output failures. The result stays a reviewable draft, no cross-provider corroboration may be implied, and every publication artifact must carry a visible notice naming the waived provider, the authorizing party and the waiver timestamp.", "Superseded-source hold: 40 CFR 1502.14 and 1505.2 are cited from the 2024 CFR annual edition and were removed effective 11 April 2025 by the CEQ interim final rule at 90 FR 10610; every rendition of the alternative-set and outcome findings must carry the historical-pattern-only label and must not present them as binding requirements.", "Relation-ledger hold: the frozen contract registers only the WM-ACT-032 inbound and WM-KNW-008 outbound relations, while the draft assigns or defers ownership to WM-REC-010, WM-ACT-024, WM-POL-021 and parent WM-KNW-007. Promotion beyond reviewable draft is blocked until those relations are registered or the dependent ownership statements are withdrawn.", "Artifact-identity hold: the serial naming rule registers only GOVLED, GOVACC and GOVDIS, leaving four serial artefacts without registered discriminators, and the dissent record scopes its sequence per dissenting party against the declared decision-master scope. Publish a complete discriminator registry and one sequence scope before any artefact name is minted.", "Independent second-provider review was explicitly waived by the repository owner; this Claude-only result remains a reviewable draft." ], "deferredResearch": [ "Re-resolve all 33 source URLs, organisations, editions and version pins against live publishers, and reconcile the DMN 1.5 dating, the Green Book 2026 edition, the CiTO 2.8.2 modification date and the NIST SP 800-53 catalog release before any conformance-adjacent statement is published.", "Obtain the paywalled normative texts — ISO/IEC/IEEE 42010-2022 and 15026-2 on rationale and assurance cases, ISO 16175-1:2020 and the ISO 23081 records-metadata series — so that the records-metadata entity model and the architecture-rationale requirements rest on the most specific standard rather than on inference from ISO 15489-1 and NARA.", "Retrieve the OECD conflict-of-interest guidance or an equivalent cross-domain source, since the participation and conflict-of-interest structure currently rests on one EU procedural instrument plus a data-protection instrument and is declared a gap.", "Search for, or explicitly refute the existence of, a normative structure for counterfactual and causal rationale and for governed reason-weighting and aggregation semantics; until then weights remain asserted values that must not be summed or compared across decisions.", "Reconcile planning/VERCY-MODEL-RELATIONS.csv with the drafted boundaries by registering the WM-KNW-007 parent relation and the WM-REC-010, WM-ACT-024 and WM-POL-021 references, or by removing the ownership assignments that depend on them.", "Publish an artefact-kind discriminator registry covering the comparison matrix, sensitivity record, dissent record and explanation rendition, and settle whether dissent-record sequences are scoped to the decision master record or to the dissenting party.", "Add a governed function that records a constituent reason and issues the statement of reasons, closing the one artefact in the model that has no producing function, and re-check function-to-artefact coverage across all 14 artefacts afterwards.", "Consult records, access and decision-governance regimes from Asia-Pacific, African, Latin American and Middle Eastern jurisdictions, none of which was consulted in this pass, before any claim of cross-jurisdictional applicability is made." ] }, "statistics": { "sources": 33, "bundles": 9, "layers": 18, "findings": 39, "questions": 175, "artifacts": 14, "functions": 21 } }