Version: 0.1
Date: 16 August 2026
Status: completed development-cycle record before adversarial architecture review
1. Register method#
Each candidate below is assessed explicitly against the sixteen-stage development cycle required by the Project Instructions. The proposed provision at step 15 is candidate text, not yet Standard text. The disposition records whether it may proceed to the adversarial architecture review as a requirement, principle, recommendation or guidance, or is deleted/open.
Evidence-strength ratings mean:
- Strong: supported by several mature primary professional sources and/or consistent authoritative practice, with a clear category application;
- Moderate: credible sources and reasoning exist, but the category-specific rule or threshold has limited direct validation;
- Limited: plausible but based mainly on analogy, narrow evidence or immature practice;
- Practice-derived: principally supported by established or Westbridge House practice;
- Hypothesis: not ready for normative use.
2. Candidate overview#
| Candidate | Short subject | Pre-adversarial disposition | Evidence strength |
|---|---|---|---|
| CAT-01 | Ex ante defensibility, not outcome warranty | Principle | Strong |
| SCOPE-01 | Truthful engagement-level conformity claim | Requirement | Moderate to strong |
| QUAL-01 | Qualification and acceptance | Requirement | Strong |
| QUEST-01 | Documented decision question | Requirement | Strong |
| MAND-01 | Mandate, access and role authority | Requirement | Strong |
| COMP-01 | Collective competence and resources | Requirement | Strong |
| IND-01 | Independence and conflicts | Requirement | Strong for core; moderate for specific prohibitions |
| EVID-01 | Adequate basis and provenance | Requirement | Strong concepts; moderate category threshold |
| REL-01 | Material third-party reliance | Requirement | Strong by analogy |
| UNC-01 | Assumptions and uncertainty | Requirement | Strong |
| CHAL-01 | Genuine independent challenge | Requirement | Strong for outcome; moderate operational threshold |
| DISS-01 | Material dissent disposition | Requirement | Moderate |
| CONC-01 | Conclusion boundaries | Requirement | Strong |
| REP-01 | Minimum report content | Requirement | Strong |
| AUTH-01 | Human output authorisation and client decision authority | Requirement | Strong role basis; moderate terminology |
| TECH-01 | Material delegated or technology-enabled analysis | Conditional requirement | Moderate |
| DOC-01 | Traceable documentation | Requirement | Strong by analogy |
| PROP-01 | Proportionality with a non-negotiable floor | Principle | Strong |
| NONC-01 | Inability to conclude, withholding and withdrawal | Requirement | Strong by analogy |
| AQT-01 | Mandatory “Accepted Question” construct | Delete; replaced by QUEST-01 | Practice-derived |
| SEP-01 | Universal separation from all implementation services | Recommendation/conditional safeguard | Limited to moderate |
| REV-01 | Mandatory separate independent quality reviewer | Recommendation for selected cases | Moderate in audit; limited for category |
| TECHNIQUE-01 | Mandatory red team, premortem or independent model | Guidance only | Moderate for some techniques; limited universal basis |
| CONF-01 | Mandatory numerical confidence score | Delete; open issue | Limited/hypothesis |
| DISS-02 | Mandatory unanimous resolution of dissent | Delete | Limited/category conflict |
| REPERF-01 | Reperform all third-party work | Delete | Contradicted by proportional reliance practice |
| ARCH-01 | Prescribed canonical case file, agents or technology stack | Delete | Westbridge House practice only |
| CERT-01 | Certification or provider-level conformity | Out of scope | Hypothesis |
CAT-01 — Ex ante professional defensibility, not outcome warranty#
Disposition: principle; reflected in scope and reporting interpretation, not a standalone conformity requirement.
Evidence strength: strong.
| Development stage | Record |
|---|---|
| 1. Problem definition | Decision outcomes are affected by uncertainty. A provider or client may market a favourable outcome as proof of quality or use conformity as a warranty; hindsight and outcome bias make that misleading. |
| 2. Scope relevance | Directly relevant to what the Standard can claim. It prevents category inflation but does not prescribe general decision governance. |
| 3. Source classification | GOV-01–03 (governance authority); DS-04–06 (hindsight/outcome bias, academic evidence); RISK-01 (uncertainty); ASR-01 (bounded engagement conclusions). |
| 4. Necessity test | Necessary as an interpretive principle. Without it, even well-drafted requirements could be used as outcome certification. |
| 5. Non-duplication test | Decision science distinguishes process from outcome; assurance defines its own conclusions. The category still needs to state the boundary because no adjacent framework automatically governs every engagement. No new outcome-quality doctrine is created. |
| 6. Normative classification | Principle, because “professional defensibility” cannot be observed as an isolated binary act. Its operational consequences appear in CONC-01, REP-01 and AUTH-01. |
| 7. Independence test | Coherent for any provider and unrelated to Westbridge House existence. |
| 8. Generalisability test | Applies across sectors and decision types. |
| 9. Observability test | Observable indirectly through report date, disclosed basis/boundaries and absence of outcome warranties. Those observations belong to other requirements. |
| 10. Proportionality test | Always applies; it does not increase work effort. |
| 11. Adversarial review | A provider might state the principle in fine print while prominently marketing “approval” or “certainty”. REP-01 and AUTH-01 must control the actual report and representations. |
| 12. Boundary review | Reduces confusion with assurance, warranties, certification and client decision approval. It does not deny that outcomes matter for later learning. |
| 13. Implementation-neutrality review | No workflow or technology prescribed. |
| 14. Terminology review | Use ex ante professional defensibility only in explanation; avoid a defined “decision quality score”. |
| 15. Drafting | Principle: Conformity concerns the professional basis, conduct and authorised output of the engagement at the report date. It does not warrant the future outcome or transfer responsibility for the consequential decision. |
| 16. Open issue record | OI-01: whether and how engagement-level conformity may be stated without implying certification or outcome validation. |
SCOPE-01 — Truthful engagement-level conformity claim#
Disposition: requirement; proceed.
Evidence strength: moderate to strong for truthful/non-misleading representation; limited evidence on future market interpretation.
| Development stage | Record |
|---|---|
| 1. Problem definition | A provider may claim “aligned”, “partly compliant” or “certified” after selecting only convenient clauses, or use the Standard to imply assurance, provider approval or outcome validation. |
| 2. Scope relevance | Directly controls how the category and Standard may be represented; it does not establish a certification scheme. |
| 3. Source classification | ETH-01 (integrity/professional behaviour); DRAFT-01 (requirements and conformity-oriented drafting conventions); ASR-01–02 (clear engagement classification); CAT-01 reasoning. No separate conformity-assessment framework is imported. |
| 4. Necessity test | Necessary if the Standard permits any conformance representation. Silence would allow selective self-labelling and pseudo-certification. |
| 5. Non-duplication test | Truthful professional representation is established elsewhere, but not every implementer is subject to one code and no source defines a claim for this category. The candidate is narrow and does not create assessment procedures. |
| 6. Normative classification | Requirement governing any voluntary claim. An engagement can use the Standard without making a conformity claim. |
| 7. Independence test | Applies equally to Westbridge House and competitors; it prevents owner privilege. |
| 8. Generalisability test | Works for internal and external engagements and any report format. Provider-level claims remain prohibited/out of scope. |
| 9. Observability test | Practitioner conformance evaluation; report wording; absence of certification/assurance claims; mapping from applicable requirements to records; treatment of departures. |
| 10. Proportionality test | All applicable requirements must be met for a claim; proportionality affects their implementation, not selective applicability. Truly conditional requirements apply only when their condition exists. |
| 11. Adversarial review | “Based on”, “inspired by” or “substantially aligned” can be used as near-conformity marketing; a clean report may hide a material departure. The Standard cannot police all speech, but it can reserve a formal conformity statement and forbid partial conformity. |
| 12. Boundary review | A self-declared engagement claim is not assurance, certification, accreditation, legal compliance or provider approval and creates no conclusion about the client. |
| 13. Implementation-neutrality review | No external assessor, mark, registry, licence or evidence platform. |
| 14. Terminology review | Use engagement was performed in conformity with... only as a self-declared engagement statement. Do not use certified, accredited, approved or assured compliant. |
| 15. Drafting | Proposed provision: A practitioner shall state that an engagement was performed in conformity with this Standard only after evaluating and documenting that all applicable requirements were satisfied. Partial conformity shall not be claimed. Where a requirement was not satisfied, the practitioner may describe the Standard as a reference used and shall describe the material departure where relevant, but shall not use wording that reasonably implies conformity. A conformity statement applies only to the identified engagement and report; it shall not be represented as certification, assurance, accreditation, provider approval, legal compliance or validation of the client decision or outcome. |
| 16. Open issue record | OI-01: user testing of claim wording and whether self-declaration should remain in v1.0; any future third-party conformity assessment is a separate out-of-scope project. |
QUAL-01 — Qualification and engagement acceptance#
Disposition: requirement; proceed.
Evidence strength: strong.
| Development stage | Record |
|---|---|
| 1. Problem definition | A sponsor can commission a second opinion too late, without a real decision, with inadequate access, to obtain comfort or to legitimise a foregone conclusion. Beginning work before testing feasibility makes later independence and scope controls ineffective. |
| 2. Scope relevance | Constitutive: the category exists only for an identifiable prospective consequential decision before material commitment. |
| 3. Source classification | ASR-01–02 (formal acceptance standards); IA-01 (objectives, scope, resources); ETH-01 (decline where threats cannot be addressed); PRACT-01–03 (gate selection and timing); GOV-01–03 (usable informed input). |
| 4. Necessity test | Necessary. Later disclosure cannot cure an engagement that was never capable of delivering a defensible and timely opinion. |
| 5. Non-duplication test | Assurance/internal-audit acceptance rules cover their own engagements; public gate schemes use prescribed eligibility. No cross-sector second-opinion qualification rule exists. The candidate adapts, rather than reproduces, those regimes. |
| 6. Normative classification | Requirement. Acceptance is an observable professional decision and a necessary condition of category use. |
| 7. Independence test | Survives without Westbridge House; used in mature professions and public review regimes. |
| 8. Generalisability test | Applies to internal and external practitioners, individual or team. Exact acceptance procedure is left open. |
| 9. Observability test | Dated acceptance/continuance record identifying the decision, commitment point, timing, intended use, preliminary scope, access, competence and independence assessment; evidence of decline/reclassification where failed. |
| 10. Proportionality test | Depth of assessment scales with the decision, but every factor must be addressed. Urgency changes evidence and conclusion boundaries; it does not remove qualification. |
| 11. Adversarial review | A checklist can be completed with unsupported “yes” answers; “next stage” can be invented after commitment. The record must contain the basis and actual remaining decision latitude. Continuance reassessment prevents initial formalism. |
| 12. Boundary review | Acceptance must not describe work as assurance unless separately qualified under ASR-01. It does not determine the client's governance process. |
| 13. Implementation-neutrality review | No intake form, scoring system or monetary threshold prescribed. |
| 14. Terminology review | Use qualification and acceptance; do not use Westbridge House qualification scores. Relevant material commitment allows staged decisions. |
| 15. Drafting | Proposed provision: Before accepting, and when material conditions change before release, the practitioner shall determine and document that: (a) an identifiable consequential decision and relevant material commitment point exist; (b) the report can be delivered while the decision authority retains meaningful ability to act on it; (c) a decision question can be bounded; (d) the intended use is legitimate and not misleading; and (e) adequate mandate, access, competence, resources and independence exist or can be obtained. If these conditions are not met, the practitioner shall decline, reclassify or discontinue the engagement and shall not represent the work as a conforming independent second opinion. |
| 16. Open issue record | OI-02: minimum information required before acceptance; OI-03: emergency and compressed-timing engagements; OI-04: what constitutes meaningful remaining decision latitude. |
QUEST-01 — Documented decision question and scope#
Disposition: requirement; proceed.
Evidence strength: strong for scope discipline; moderate for exact category contents.
| Development stage | Record |
|---|---|
| 1. Problem definition | A broad or sponsor-framed question can make the work unfalsifiable, hide excluded alternatives, shift during the engagement or permit a conclusion irrelevant to the actual commitment. |
| 2. Scope relevance | Directly defines the unit of the second-opinion engagement. It is not a general requirement for all board papers. |
| 3. Source classification | IA-01 Standards 13.3–13.4 (formal professional standards); ASR-01–02 (engagement terms); CONS-01 (guidance); PRACT-02–03 (decision-gate practice); WH-01 (practice input only). |
| 4. Necessity test | Necessary. Independence and evidence sufficiency cannot be evaluated without knowing the decision question and permissible conclusion. |
| 5. Non-duplication test | Objectives and scope already exist elsewhere. The new content is limited to the decision connection: actual decision, commitment point, authority, intended use and conclusion. The Accepted Question label/template is not duplicated. |
| 6. Normative classification | Requirement. A bounded documented question is observable and constitutive. |
| 7. Independence test | The construct remains valid without Westbridge House when expressed generically. |
| 8. Generalisability test | Works for transactions, capital programmes, policy, technology or strategic commitments; form and level of detail vary. |
| 9. Observability test | Approved question/scope record and change history; the report restates the current decision question and material exclusions. |
| 10. Proportionality test | Detail scales with complexity. The minimum elements are always required. |
| 11. Adversarial review | A sponsor may word the question narrowly (“is the model arithmetically correct?”) while presenting the output as a decision endorsement. Qualification, intended use and material-exclusion disclosure prevent that. |
| 12. Boundary review | The question must not promise legal advice, audit or assurance unless those components are separately governed. It does not give the practitioner the client's decision mandate. |
| 13. Implementation-neutrality review | No prescribed template, “canonical question” workflow or sign-off system. |
| 14. Terminology review | Decision question is adopted. Accepted Question is rejected as a generic term; engagement question is less decision-specific. |
| 15. Drafting | Proposed provision: Before substantive work, the practitioner and engaging party shall document a decision question that identifies: (a) the consequential decision and relevant commitment point; (b) the decision authority; (c) the purpose, intended users and intended use; (d) the decision basis, period and material matters within scope; (e) known material exclusions and constraints; (f) the type and boundaries of conclusion the practitioner may express; and (g) the required report timing. Material changes shall be documented, reassessed under QUAL-01 and reflected in the report. |
| 16. Open issue record | OI-05: whether intended users must consent; OI-06: scope treatment for legally privileged or inaccessible information; OI-07: minimum change-control threshold. |
MAND-01 — Mandate, access and role authority#
Disposition: requirement; proceed and likely merge with QUEST-01 in the final Standard.
Evidence strength: strong.
| Development stage | Record |
|---|---|
| 1. Problem definition | A practitioner may be nominally independent but lack access, the right to challenge, freedom to communicate adverse findings or clarity about responsibilities. Conversely, the report may be treated as decision approval. |
| 2. Scope relevance | Constitutive to a professional engagement; confined to the authority needed for this work. |
| 3. Source classification | IA-01 (charter, scope, access, communication); ASR-01–02 (engagement terms and roles); GOV-01–03 and GOV-05 (decision responsibility); PRACT-01–03 (review informs, authority decides); CONS-01. |
| 4. Necessity test | Necessary. Independence without operative authority or access is formal only; authority without limits creates governance substitution. |
| 5. Non-duplication test | Contracts and charters already address mandates, but no universal contract applies. The candidate states only the functional professional minimum and leaves legal terms to the engagement. |
| 6. Normative classification | Requirement. Adequacy of mandate and role separation is a conformance condition. |
| 7. Independence test | General professional principle, not Westbridge House-dependent. |
| 8. Generalisability test | Can be implemented through contract, charter, board resolution, terms of reference or equivalent. |
| 9. Observability test | Written mandate/terms; evidence of access requests and responses; unaltered practitioner conclusion; report statement of decision authority and limitations. |
| 10. Proportionality test | Formality varies, but freedom to challenge/report and the decision-authority boundary do not. |
| 11. Adversarial review | Terms may promise “full access” while sponsors control interviewees or omit known repositories. Access limitations must feed EVID-01, CONC-01 and NONC-01. A boilerplate disclaimer cannot cure actual practitioner decision-making. |
| 12. Boundary review | Separates the engagement from management responsibility, board approval, audit authority and legal compulsion. |
| 13. Implementation-neutrality review | No specified contract, data room, interview process or reporting line. |
| 14. Terminology review | Use mandate, engaging party, decision authority and practitioner. Avoid accountable principal and examiner powers. |
| 15. Drafting | Proposed provision: The practitioner shall have and document a mandate adequate to obtain and assess relevant information, contact relevant sources within agreed legal and confidentiality limits, challenge the decision basis, determine necessary work, and communicate an adverse, qualified or no conclusion without sponsor alteration. The mandate shall distinguish the engaging party, intended users, information providers, output authoriser and decision authority and shall state that the practitioner does not make or approve the consequential decision. Material access or authority restrictions shall be treated under CONC-01 and NONC-01. |
| 16. Open issue record | OI-06: privileged/inaccessible information; OI-08: internal practitioner reporting lines; OI-09: distribution and reliance language across jurisdictions. |
COMP-01 — Collective competence and resources#
Disposition: requirement; proceed.
Evidence strength: strong.
| Development stage | Record |
|---|---|
| 1. Problem definition | An independent but technically or methodologically unqualified practitioner can give an authoritative-sounding opinion, fail to recognise specialist limits or delegate without understanding. |
| 2. Scope relevance | Directly relevant to professional defensibility. It does not create a general occupational regime. |
| 3. Source classification | IA-01 Standards 3.1, 4.2 and 13.5; ASR-01; ASR-04; ETH-01; DS-09 (limits of confidence); PRACT-01–04. Formal standards strongly support the core. |
| 4. Necessity test | Necessary. Independence does not compensate for lack of competence, and specialist reliance does not transfer practitioner responsibility. |
| 5. Non-duplication test | Competence is covered well in existing professions. A concise functional requirement is still necessary because the category is multidisciplinary and open to providers outside those professions. No proprietary competency model or credential is added. |
| 6. Normative classification | Requirement. Engagement-specific competence and resources can be evidenced. |
| 7. Independence test | Fully independent of Westbridge House and consistent across professions. |
| 8. Generalisability test | Collective competence permits different team sizes and specialists; regulated work remains subject to applicable licensing. |
| 9. Observability test | Competence/resource assessment mapped to material question areas; team roles; specialist rationale; evidence that output authoriser understands material judgments; documented limitations/referral. |
| 10. Proportionality test | Breadth and depth scale with subject matter, novelty, complexity and stakes. Minimum competence cannot be scaled below the work actually undertaken. |
| 11. Adversarial review | Credential lists can substitute for actual question competence; a famous expert may supply a name without reviewing the conclusion. Observability must link competence and participation to material work. |
| 12. Boundary review | The Standard does not authorise regulated legal, audit, medical, engineering or investment opinions. Reliance on licensed experts is addressed by REL-01. |
| 13. Implementation-neutrality review | No degree, certification, team composition or staffing ratio required. |
| 14. Terminology review | Use collective competence and capability/resources in ordinary professional senses. Avoid creating a “certified examiner”. |
| 15. Drafting | Proposed provision: Before and throughout the engagement, the practitioner shall determine and document that the individuals performing and authorising the work collectively possess the subject-matter competence, critical-evaluation capability, experience, time and resources necessary for the decision question. Where material matters require expertise not held by the practitioner team, competent specialist input shall be obtained or the scope and conclusion shall be limited. The practitioner shall not independently conclude on a reserved or specialist matter beyond its competence or authority. |
| 16. Open issue record | OI-10: sector-specific competence indicators; OI-11: competence of the output authoriser in multidisciplinary cases; future practice validation. |
IND-01 — Independence and conflicts of interest#
Disposition: requirement; proceed, with proportional safeguards but absolute treatment of conclusion-contingent remuneration and unmanageable material interests.
Evidence strength: strong for threat-based assessment; moderate for category-specific absolute guardrails.
| Development stage | Record |
|---|---|
| 1. Problem definition | A provider can appear separate while incentives, prior authorship, implementation revenue, personal interests, client dependence or pressure impair objective challenge. Disclosure can be used as a substitute for a real response. |
| 2. Scope relevance | Independence is constitutive of the named category; without it the work may still be advice or review, but not an independent second opinion. |
| 3. Source classification | ETH-01 (formal ethics/independence conceptual framework); IA-01 Standards 2.1–2.3; ASR-01; ASR-07; GOV-01–04; IA-02; WH-01 as practice input. |
| 4. Necessity test | Necessary. The category's value depends on freedom to test and contradict the sponsor's position. |
| 5. Non-duplication test | Existing threat taxonomies are mature and should not be renamed. Many potential practitioners, however, are not subject to IESBA or IIA rules, and those rules do not define independence from this decision. A concise category minimum is required. |
| 6. Normative classification | Requirement. Threat assessment, responses and disposition are observable; independence itself requires professional judgment. |
| 7. Independence test | The rule is grounded externally and remains coherent for any provider. No Westbridge House structural model is required. |
| 8. Generalisability test | Applies to internal and external practitioners. Safeguards may include role removal, separate personnel/governance, independent review, fee changes or withdrawal. |
| 9. Observability test | Dated conflict/independence assessment covering financial interests, fees, relationships, prior roles/work, implementation interests, advocacy, familiarity, intimidation and influence; recorded safeguards; disclosures; decline/withdrawal decisions; evidence of adverse-conclusion freedom. |
| 10. Proportionality test | Extent of inquiry and safeguards scales with threat severity. A material direct interest or conclusion-contingent remuneration cannot be made acceptable merely by a smaller engagement or disclosure. |
| 11. Adversarial review | Practitioners may declare “no conflict” narrowly, disclose a lucrative implementation opportunity in fine print, or use a separate legal entity with common incentives. The test concerns substance and reasonable appearance, not organisation charts. Client consent alone is not conclusive. |
| 12. Boundary review | Stricter applicable independence rules prevail. The requirement should not imply the full IESBA International Independence Standards apply to a non-assurance engagement. It also avoids claiming that external status guarantees independence. |
| 13. Implementation-neutrality review | No required firm structure, information barrier, external provider or reviewer model. Universal implementation separation is separately tested in SEP-01. |
| 14. Terminology review | Use actual, potential or perceived threat/impairment, objective judgment and reasonable and informed observer; these are internationally recognisable. Define independence in category terms. |
| 15. Drafting | Proposed provision: Before acceptance and throughout the engagement, the practitioner shall identify, evaluate and document interests, relationships, prior roles, services, remuneration arrangements and influence that create actual, potential or perceived threats to objective judgment, genuine challenge or freedom to reach and communicate an adverse, qualified or no conclusion. The practitioner shall eliminate the circumstance, apply safeguards that reduce the threat to an acceptable level, or decline/withdraw. Disclosure or client consent alone shall not make an otherwise unacceptable threat acceptable. The practitioner shall not accept remuneration contingent on the content of the conclusion or the client's decision outcome and shall not continue while the practitioner or output authoriser holds an unmanaged material direct financial interest in that outcome. Stricter applicable independence rules shall prevail. |
| 16. Open issue record | OI-12: prior authorship and cooling-off periods; OI-13: implementation-service interests; OI-14: organisational independence for internal functions; OI-15: reasonable-observer formulation across legal systems. |
EVID-01 — Adequate basis, evidence quality and provenance#
Disposition: requirement; proceed.
Evidence strength: strong for evidence qualities and skepticism; moderate for the category-specific “adequate basis” threshold.
| Development stage | Record |
|---|---|
| 1. Problem definition | A professional-looking conclusion can be based on a document dump, untraceable claims, selective evidence, stale data, sponsor assertions or analyst-generated inferences presented as facts. Prospective matters also cannot always be verified as historical facts. |
| 2. Scope relevance | Central to whether the second opinion is professionally defensible; limited to material aspects of the decision question. |
| 3. Source classification | IA-01 Standards 4.3, 14.1 and 14.6; ASR-01 and ASR-03; ASR-08; DD-02–03; PRACT-04; DS-08–09. Mostly formal professional standards and government method, applied analogically. |
| 4. Necessity test | Necessary. Without a conclusion-relative evidence threshold, challenge can be rhetorical rather than professional. |
| 5. Non-duplication test | Audit/assurance evidence concepts are mature but carry criteria, assertions and assurance objectives. The candidate adopts relevance, reliability and provenance while defining a non-assurance “adequate basis” matched to the stated conclusion. |
| 6. Normative classification | Requirement. Workpapers and report provide observable evidence; adequacy remains professional judgment. |
| 7. Independence test | Fully independent of Westbridge House; based on multiple external frameworks. |
| 8. Generalisability test | Applies to qualitative and quantitative decisions. It permits facts, expert judgments, forecasts, models and reasoned inference, provided their status and limits are clear. |
| 9. Observability test | Source register or equivalent; provenance and currency for material information; separation of source claim, verified fact, assumption and practitioner inference; treatment of inconsistent/missing information; link from material conclusions to support. |
| 10. Proportionality test | Extent varies with materiality, uncertainty, reliability, complexity and conclusion strength. A strong conclusion requires a stronger basis. Proportionality cannot convert an inadequate basis into an unqualified conclusion. |
| 11. Adversarial review | A provider may count sources rather than assess quality, cite many mutually dependent sources as corroboration, or treat management representations as verified. The requirement focuses on fitness, independence of corroboration and status, not volume. |
| 12. Boundary review | Avoid the unexplained phrase sufficient appropriate evidence and any assurance level. The second opinion does not warrant completeness or audit all inputs. Legal/regulated evidence duties remain separate. |
| 13. Implementation-neutrality review | No prescribed database, source register schema, sampling method or canonical case file. |
| 14. Terminology review | Define adequate basis for the conclusion. Use relevant, reliable, provenance, corroboration and inconsistency in ordinary professional senses. |
| 15. Drafting | Proposed provision: The practitioner shall plan and perform work sufficient to establish an adequate basis for the stated conclusion and its boundaries. For material information, the practitioner shall assess relevance, reliability, provenance, currency and consistency; distinguish source claims, established facts, assumptions, estimates and practitioner inferences; and address significant missing, conflicting or doubtful information. The extent of work shall respond to the materiality of the matter, the quality of available support, uncertainty and the strength of conclusion proposed. The practitioner shall not imply exhaustive verification or assurance that was not performed. |
| 16. Open issue record | OI-16: validation of the adequate-basis threshold; OI-17: use of confidential/unshareable evidence; OI-18: source quality for novel markets or technologies. |
REL-01 — Material third-party reliance#
Disposition: requirement; proceed.
Evidence strength: strong by analogy; moderate across all types of third-party work.
| Development stage | Record |
|---|---|
| 1. Problem definition | The practitioner may rely on legal, technical, commercial, financial, expert or model work whose purpose, scope, date, assumptions, competence, objectivity or use rights do not support the second-opinion conclusion. A citation can conceal unsupported reliance. |
| 2. Scope relevance | Directly relevant whenever third-party work materially supports the conclusion; not a general vendor-management standard. |
| 3. Source classification | ASR-04 (formal expert standard); ETH-02 (future-effective professional direction); IA-01; DD-02–03; ASR-01; PRACT-04. |
| 4. Necessity test | Necessary as a conditional requirement. Multidisciplinary decision work often cannot be re-performed by one practitioner, so defensibility depends on controlled reliance. |
| 5. Non-duplication test | Expert reliance is mature in audit, but those procedures apply to defined expert relationships. Due diligence covers third-party reports but not retained responsibility for an independent decision conclusion. The candidate integrates different reliance sources at a functional level. |
| 6. Normative classification | Requirement when reliance is material. Immaterial references do not trigger full assessment. |
| 7. Independence test | Survives without Westbridge House and mirrors established professional responsibility. |
| 8. Generalisability test | Applies to the practitioner's experts, client experts, prior diligence, external professional opinions, datasets, valuation work and other material analysis. Technology-generated work also engages TECH-01. |
| 9. Observability test | Reliance inventory; assessment of purpose/scope/date/method and relevant competence/capability/objectivity; permissions or noted restrictions; evaluation of key assumptions and fit; additional work; report disclosure of material reliance and limitations. |
| 10. Proportionality test | Assessment scales with reliance, materiality, complexity and source risk. It need not re-perform all work. A dispositive expert input requires deeper evaluation than background context. |
| 11. Adversarial review | A provider may label work “background only” while the conclusion depends on it, rely on a prestigious firm name, or confuse legal permission to receive a report with evidential fitness. Material dependency, not the label, triggers the requirement. |
| 12. Boundary review | The practitioner does not convert another professional's work into its own legal or assurance opinion. Applicable reliance restrictions and professional reservations must be respected and reported where material. |
| 13. Implementation-neutrality review | No approved-vendor list, credential, expert form or required re-performance. |
| 14. Terminology review | Use third-party work broadly and material reliance for dependence. Avoid implying legal third-party reliance rights merely from evidential use. |
| 15. Drafting | Proposed provision: Where the conclusion materially relies on work, information or judgment produced outside the practitioner team, the practitioner shall evaluate its purpose, scope, date, method, material assumptions and limitations and, to the extent relevant, the provider's competence, capability and objectivity. The practitioner shall determine whether the work is fit to support the particular use, perform or obtain additional work where necessary, and disclose material reliance and unresolved limitations in the report. Permission to access or cite work shall not by itself establish its fitness. The practitioner retains responsibility for its own conclusion. |
| 16. Open issue record | OI-09: legal reliance/distribution terms; OI-19: how far to assess provider objectivity; OI-20: proprietary expert/model methods; OI-21: use of another assurance report. |
UNC-01 — Material assumptions and uncertainty#
Disposition: requirement; proceed.
Evidence strength: strong for transparency and restraint; moderate for operational depth across decision types.
| Development stage | Record |
|---|---|
| 1. Problem definition | Forecasts, plans and valuation narratives can hide material assumptions as facts, suppress ranges, ignore dependencies or present precision unsupported by information. Missing evidence can be silently replaced by optimistic judgment. |
| 2. Scope relevance | Prospective consequential decisions are inherently uncertain; this is not an enterprise risk-management requirement. |
| 3. Source classification | RISK-01–02; PRACT-04; DD-02; ASR-01; DS-08–10; DS-04–06. Guidance, professional standards and academic evidence converge on explicit uncertainty and ex ante reasoning. |
| 4. Necessity test | Necessary. An evidence-relative opinion is misleading if its material assumptions and uncertainty do not affect analysis and reporting. |
| 5. Non-duplication test | Risk frameworks already define and manage uncertainty, but not how it bounds a second-opinion conclusion. The candidate requires linkage, not a new risk taxonomy or risk appetite process. |
| 6. Normative classification | Requirement. Material assumptions/uncertainties and their treatment can be observed. Specific analytic techniques remain recommendations/guidance. |
| 7. Independence test | Supported independently and applicable without Westbridge House. |
| 8. Generalisability test | Applies to qualitative unknowns and quantified estimates. It does not assume reliable probabilities or a model are available. |
| 9. Observability test | Assumption/uncertainty record or equivalent; source and rationale; testing of materially outcome-sensitive items; identification of missing information and dependencies; corresponding report boundaries. |
| 10. Proportionality test | Depth depends on sensitivity, uncertainty, reversibility and decision significance. Formal modelling is required only when necessary for an adequate basis. |
| 11. Adversarial review | A long generic risk list can conceal the few assumptions that drive the decision; unexplained ranges can create false sophistication. The test is material effect on decision/conclusion, not list length or numerical form. |
| 12. Boundary review | The practitioner does not set client risk appetite or provide a probabilistic assurance level. Legal, actuarial or valuation uncertainty may require specialists. |
| 13. Implementation-neutrality review | No mandatory scenario tool, Monte Carlo model, confidence scale, assumption register or sensitivity format. |
| 14. Terminology review | Use material assumption, uncertainty, estimate, scenario and sensitivity. Do not use confidence boundary or confidence score. |
| 15. Drafting | Proposed provision: The practitioner shall identify and assess assumptions, estimates, dependencies, information gaps and uncertainties that could materially affect the decision basis or conclusion. The practitioner shall test their plausibility and decision effect to the extent necessary and proportionate, distinguish them from established facts, and reflect material sensitivity, unresolved uncertainty and missing information in the conclusion boundaries and report. Quantification shall be used only where the subject and available information support it. |
| 16. Open issue record | OI-22: deep uncertainty; OI-23: thresholds for sensitivity analysis; OI-24: sector-specific quantitative expectations; CONF-01 separately tests numerical confidence. |
CHAL-01 — Genuine independent decision challenge#
Disposition: requirement; proceed as a performance requirement.
Evidence strength: strong for the need to challenge and consider contrary information; moderate for a universal stopping threshold.
| Development stage | Record |
|---|---|
| 1. Problem definition | A provider can call a summary, validation of arithmetic or polite discussion a “second opinion” while accepting the sponsor's frame, preferred option and evidence selection. Confirmation, escalation and shared-information effects can reinforce this failure. |
| 2. Scope relevance | Constitutive distinction from information collation, facilitation and ordinary diligence. It is limited to material aspects of the decision question. |
| 3. Source classification | GOV-02–03 (governance code/guidance); IA-01 Standard 4.3; DS-01–03, DS-07–08, DS-10–12 (academic evidence); PRACT-01–04 (established review practice). |
| 4. Necessity test | Necessary. Without active disconfirmation, “independent” describes only organisational position, not work performed. |
| 5. Non-duplication test | Constructive challenge and professional skepticism exist, but no adjacent standard ties them to a bounded cross-sector pre-commitment decision conclusion. The candidate defines an outcome, not a rival skepticism doctrine. |
| 6. Normative classification | Requirement. Genuine performance must be evident in work and reasoning; named techniques are guidance. |
| 7. Independence test | Strongly supported outside Westbridge House. |
| 8. Generalisability test | Can be achieved through interviews, independent analysis, alternative framing, reference data, red teaming, peer challenge or other methods. Small engagements need not use a separate team. |
| 9. Observability test | Record of material framing/claim tests; source selection beyond sponsor materials where reasonably available; plausible competing interpretations; counterevidence sought/received; disposition and effect on conclusion. |
| 10. Proportionality test | Search and testing scale with materiality, uncertainty, novelty and time. Challenge must remain genuine; trivial objections do not demonstrate conformity. Stopping is justified when further work is unlikely to change a proportionately bounded conclusion. |
| 11. Adversarial review | A hostile implementer can manufacture a “challenge log” of immaterial questions, appoint a friendly red team or reject all alternatives as implausible. Observability must focus on issues capable of changing the conclusion and the reasoned disposition. |
| 12. Boundary review | Challenge does not make the practitioner the decision maker, advocate or manager. It does not require exhaustive due diligence or proof that no counterevidence exists. |
| 13. Implementation-neutrality review | No required red team, premortem, independent model, agent, workshop, checklist or number of alternatives. |
| 14. Terminology review | Use genuine challenge, plausible competing interpretation and material counterevidence. Avoid “devil's advocate” as a normative role. |
| 15. Drafting | Proposed provision: The practitioner shall perform genuine independent challenge of the decision basis. The work shall, to the extent material to the decision question: (a) test the framing, objectives, feasible alternatives and material claims; (b) question material assumptions and dependencies; (c) seek and consider information reasonably capable of disconfirming the working view; (d) evaluate plausible competing interpretations; and (e) resolve or transparently disposition material counterevidence. The practitioner shall determine and document methods and extent proportionate to the decision and shall not begin with an obligation to confirm the sponsor's preferred conclusion. |
| 16. Open issue record | OI-25: stopping rule and minimum evidence of genuine challenge; OI-26: access to unshared information; OI-27: measurement without gaming; TECHNIQUE-01 tests named methods. |
DISS-01 — Material dissent disposition#
Disposition: requirement; proceed to adversarial review with a narrow definition and explicit options other than forced consensus.
Evidence strength: moderate.
| Development stage | Record |
|---|---|
| 1. Problem definition | A practitioner team or formal reviewer may hold a reasoned disagreement capable of changing the conclusion, boundaries or use. Sponsors or leaders may suppress it, or a dissenter may use an immaterial objection to block output indefinitely. |
| 2. Scope relevance | Relevant to multidisciplinary professional output and authorisation, not to every client disagreement or board debate. |
| 3. Source classification | ASR-05–07 (formal audit quality/differences standards); IA-01 Standard 13.1; GOV-02–03; DS-07 and DS-12; WH-01 as practice input. External sources support serious treatment, but category-specific disclosure is not directly validated. |
| 4. Necessity test | A disposition rule is necessary because false consensus can make the report misleading. A universal unanimity rule is not necessary. |
| 5. Non-duplication test | Audit standards generally require resolution before report dating; internal audit may communicate both positions. Neither directly answers a non-assurance multidisciplinary opinion. The candidate creates a transparent disposition, not an audit resolution rule. |
| 6. Normative classification | Requirement, but only for narrowly defined material dissent. Evidence strength and threshold require open-issue disclosure. |
| 7. Independence test | The problem and rule survive without Westbridge House; terminology was refined independently. |
| 8. Generalisability test | Applies whether dissent arises within a team, from a formally engaged specialist or in an established independent review function. A solo practitioner has no internal dissent but may still face contrary external evidence under CHAL-01. |
| 9. Observability test | Protected/available route to record reasoned disagreement; issue, rationale, evidence and impact; consultation/review; outcome (resolved, accepted, disclosed, scope/conclusion modified, withheld or withdrawal); output-authoriser decision. |
| 10. Proportionality test | Only disagreement reasonably capable of materially affecting conclusion, boundaries, authorisation or user understanding triggers the rule. Severity determines review and disclosure. |
| 11. Adversarial review | Leaders may define all dissent as comments and erase it; a dissenter may strategically claim materiality; disclosure may become an unhelpful appendix dump. The output authoriser must make and document a reasoned materiality/disposition judgment, subject to independent review where proportionate. |
| 12. Boundary review | Does not require publication of privileged, personal or legally prohibited content; if material substance cannot be communicated sufficiently, output may need to be withheld. Does not transfer report responsibility to the dissenter. |
| 13. Implementation-neutrality review | No mandatory voting, unanimity, red-team structure, ombudsperson or separate panel. |
| 14. Terminology review | Retain material dissent with a narrow defined meaning. Distinguish from management disagreement, unresolved evidence and editorial comments. |
| 15. Drafting | Proposed provision: The practitioner shall provide a means for practitioner-team members and formal reviewers to raise reasoned professional disagreement without suppression. Material dissent shall be identified, documented, evaluated and brought to the output authoriser before release. The practitioner shall record whether and how it was resolved and its effect on the work and conclusion. Unresolved material dissent that could reasonably affect an intended user's understanding of the conclusion or its boundaries shall be communicated in substance in the report; if lawful and sufficiently clear communication is not possible, the output shall be withheld or the engagement withdrawn. Dissent shall not be omitted solely to present consensus. |
| 16. Open issue record | OI-28: disclosure versus release-blocking threshold; OI-29: confidentiality/privilege; OI-30: protection and independence of dissent escalation; OI-31: solo-practitioner application. |
CONC-01 — Evidence-relative conclusion and conclusion boundaries#
Disposition: requirement; proceed.
Evidence strength: strong for bounded reporting; moderate for category-specific terminology.
| Development stage | Record |
|---|---|
| 1. Problem definition | A conclusion can sound broader, more certain or more durable than the work supports; omit a material limitation; turn conditional analysis into apparent approval; or use “confidence” language that implies assurance or calibrated probability. |
| 2. Scope relevance | Directly governs the defining professional output. |
| 3. Source classification | ASR-01 (formal assurance analogue); IA-01 Standard 15.1; ASR-02; DD-02–03; RISK-01; DS-04–06 and DS-09; PRACT-01–03. |
| 4. Necessity test | Necessary. Even good work becomes misleading if the expressed conclusion exceeds its basis or conceals its boundaries. |
| 5. Non-duplication test | Modified conclusions and limitations exist in assurance/audit; due diligence also reports limitations. This category lacks suitable common criteria and assurance levels, so it needs a restrained non-assurance conclusion rule rather than those report forms. |
| 6. Normative classification | Requirement. Response to the question and disclosed boundaries are directly observable. |
| 7. Independence test | General professional discipline, not a Westbridge House-specific “confidence boundary” construct. |
| 8. Generalisability test | Allows narrative, conditional, adverse and no-conclusion outputs across sectors. No prescribed rating or recommendation form. |
| 9. Observability test | Trace from question and evidence to conclusion; report includes point in time, scope, material assumptions/uncertainties/reliance/limitations/conditions; clear distinction between findings, inference and conclusion; no unsupported precision. |
| 10. Proportionality test | Amount of detail scales with what an intended user needs, but every material boundary is disclosed. A less material decision does not permit a false or overbroad conclusion. |
| 11. Adversarial review | A provider may bury fatal limitations after a prominent favourable headline, use “within scope” to exclude the decision's core issue, or call uncertainty “low confidence” without a method. The conclusion and boundaries must be read together and material limitations receive appropriate prominence. |
| 12. Boundary review | The provision expressly avoids reasonable/limited assurance, statistical confidence and decision approval. A separate legal or assurance conclusion must be clearly distinguished. |
| 13. Implementation-neutrality review | No mandated rating scale, traffic light, memo structure, confidence score or recommendation label. |
| 14. Terminology review | Adopt conclusion boundaries: scope, time, assumptions, uncertainty, reliance, limitations and conditions. Do not define confidence boundaries. Avoid unqualified opinion because of audit connotations. |
| 15. Drafting | Proposed provision: The practitioner shall express a conclusion that responds to the decision question and is no more definitive or extensive than its adequate basis permits. The conclusion and its presentation shall identify, with prominence proportionate to their effect, the scope and point in time, material assumptions and uncertainty, material third-party reliance, limitations, conditions and matters that could reasonably change the conclusion. Findings, source positions, practitioner inferences and the conclusion shall not be presented as interchangeable. The report shall not state or imply assurance, calibrated confidence, decision approval or an outcome warranty unless a separate applicable engagement validly provides it. |
| 16. Open issue record | OI-32: conclusion taxonomy; OI-33: prominence tests in short/digital outputs; OI-34: updating after report date; OI-35: use of recommendations versus conclusions. |
REP-01 — Minimum second-opinion report content#
Disposition: requirement; proceed.
Evidence strength: strong.
| Development stage | Record |
|---|---|
| 1. Problem definition | Users may receive a conclusion without knowing the exact decision, timing, scope, basis, reliance, limitations, dissent, authoriser or authority boundary. A persuasive slide or oral statement can become detached from the controlled professional output. |
| 2. Scope relevance | Directly defines the minimum communication necessary for the category; does not regulate all work products. |
| 3. Source classification | IA-01 Standard 15.1; ASR-01–02; DD-02–03; PRACT-01–03; GOV-02–03; DRAFT-01. |
| 4. Necessity test | Necessary. A user cannot interpret or responsibly use the conclusion without minimum context. |
| 5. Non-duplication test | Other report standards use forms designed for audit, assurance, internal audit or factual findings. This candidate selects only content necessary to the second-opinion category and does not reproduce those formats. |
| 6. Normative classification | Requirement. Presence, clarity and consistency of minimum content are observable. |
| 7. Independence test | No reliance on a Westbridge House Second Opinion Memorandum. |
| 8. Generalisability test | A report may be a memorandum, letter, secure digital output or other durable form. Material oral communications require incorporation or reference in the controlled record. |
| 9. Observability test | Authorised dated report with enumerated content; version control; consistency with the decision question and work record; identified authoriser and decision authority. |
| 10. Proportionality test | Length and presentation scale, not the presence of material information. References/annexes may reduce repetition if accessible to intended users. |
| 11. Adversarial review | Boilerplate can technically contain every item while obscuring the actual limitation; a “draft” can be used as the real opinion to avoid authorisation. Material information must be clear and the status/version unmistakable. |
| 12. Boundary review | Report must distinguish any separately governed assurance, audit, legal or specialist component. Distribution/reliance rights remain legal/contractual and are not created by the Standard. |
| 13. Implementation-neutrality review | No branded memorandum, page count, typography, portal or fixed section sequence. |
| 14. Terminology review | Use second-opinion report, output authoriser, decision authority and conclusion boundaries. |
| 15. Drafting | Proposed provision: The practitioner shall issue the authorised professional output in a durable form clearly identified as a second-opinion report. The report shall state: (a) the decision question, relevant commitment point, decision authority, intended users/use and report date; (b) scope, material exclusions and principal work performed; (c) the conclusion and a concise explanation of its basis; (d) material assumptions, uncertainty, third-party reliance, limitations and conditions; (e) unresolved material dissent required by DISS-01; (f) the identity or accountable role of the output authoriser; (g) that the report informs but does not make or approve the consequential decision; and (h) any engagement-level conformity statement permitted by SCOPE-01. The report shall be accurate, balanced, clear and not misleading when read as a whole. |
| 16. Open issue record | OI-01: conformity claims; OI-09: legal distribution/reliance; OI-33: prominence/digital reports; OI-36: naming individuals versus roles. |
AUTH-01 — Meaningful human output authorisation and client decision authority#
Disposition: requirement; proceed.
Evidence strength: strong for retained human professional and governance responsibility; moderate for generic role terminology.
| Development stage | Record |
|---|---|
| 1. Problem definition | A report can be released by workflow, management, a model or a ceremonial reviewer without a competent professional taking responsibility. Conversely, the professional sign-off can be portrayed as approval of the client's decision. |
| 2. Scope relevance | Constitutive to the professional output and its use; does not dictate the client's board process. |
| 3. Source classification | ASR-05–07 (engagement responsibility and quality review); IA-01 Standard 15.1; GOV-01–05; IA-02; PRACT-01–03; AI-01–04. |
| 4. Necessity test | Necessary. Delegated analysis requires an accountable authoriser, while governance requires that decision responsibility remain with the authorised client body. |
| 5. Non-duplication test | Existing professions use engagement partner, CAE or reviewer roles; these titles do not generalise. Governance already retains board responsibility. The category-specific gap is the explicit dual-authority relationship. |
| 6. Normative classification | Requirement. Authorisation, reviewer competence, review activity and authority statements are observable. |
| 7. Independence test | Remains necessary for any provider and is not tied to a Westbridge House reviewer gate. |
| 8. Generalisability test | The output authoriser may be the lead practitioner, engagement partner, designated internal officer or equivalent. A separate person is not universally required. |
| 9. Observability test | Named/identifiable competent human; documented review of material judgments, limitations, dissent and delegated work; evidence of authority to require changes or withhold release; dated approval; report identifies decision authority and role boundary. |
| 10. Proportionality test | Review depth scales with complexity and risk, but cannot be nominal. Human authorisation and client authority separation always apply. |
| 11. Adversarial review | A rubber-stamp reviewer can sign without access or time; a committee can diffuse responsibility; the client may issue a practitioner draft as approval. The authoriser must understand material judgments and control release of the professional output. |
| 12. Boundary review | A quality reviewer does not take over engagement work; the output authoriser does not take over the client's legal/managerial decision. Applicable governance and professional rules prevail. |
| 13. Implementation-neutrality review | No prescribed title, committee, two-signature rule, employment model or reviewer workflow. |
| 14. Terminology review | Define output authorisation and output authoriser functionally. Use decision authority for the client body/person. Avoid “professional-output gate” in public text. |
| 15. Drafting | Proposed provision: Before release, a competent and sufficiently independent human output authoriser shall perform and document review sufficient to understand the decision question, material work and judgments, conclusion boundaries, material reliance, delegated contributions and material dissent; require necessary correction or additional work; and take responsibility for authorising or withholding the professional output. The report shall identify the output authoriser or accountable role separately from the decision authority. Output authorisation shall not be represented as making, approving or guaranteeing the consequential decision, which remains with the decision authority. |
| 16. Open issue record | OI-11: multidisciplinary competence of authoriser; OI-36: name versus role; OI-37: internal function hierarchy; REV-01: when separate review is warranted. |
TECH-01 — Material delegated or technology-enabled analysis#
Disposition: conditional requirement; proceed, drafted technology-neutrally.
Evidence strength: moderate.
| Development stage | Record |
|---|---|
| 1. Problem definition | Models, AI systems, agents, external analysts or automated tools can materially shape a conclusion while obscuring input provenance, limitations, error, bias and responsibility. A human signature can be nominal rather than meaningful review. |
| 2. Scope relevance | Relevant only when delegated or technology-enabled work materially contributes to the engagement. It is not a general AI governance standard. |
| 3. Source classification | AI-01 (academic automation-bias evidence); AI-02 (binding law where applicable); AI-03–04 (authoritative voluntary principles); IA-01; ASR-04–08. |
| 4. Necessity test | Conditional requirement is necessary because ordinary evidence and authorisation rules may not make the role and limitations of opaque or delegated analysis visible. Tool-specific governance is unnecessary. |
| 5. Non-duplication test | Applicable AI law and organisational governance remain controlling. The candidate only ensures that delegation cannot bypass EVID-01, REL-01, DOC-01 or AUTH-01 within this engagement. |
| 6. Normative classification | Requirement when contribution is material. Non-material productivity tools need not be inventoried. |
| 7. Independence test | The rule survives any particular Westbridge House agent/MCP/A2A implementation and applies equally to non-AI models and delegated analysts. |
| 8. Generalisability test | Covers internal models, spreadsheets, AI, agents, outsourced analysis and other systems, using controls appropriate to their role and risk. |
| 9. Observability test | Record of material tool/delegate role, inputs and source provenance, material assumptions/limitations, validation or corroboration, changes/versions where relevant, human review and override; report disclosure where needed to understand conclusion boundaries. |
| 10. Proportionality test | Controls scale with materiality, opacity, autonomy, novelty, error consequences and availability of independent checks. The human remains accountable in all material cases. |
| 11. Adversarial review | A provider can call an AI-generated conclusion “drafting assistance”, flood the file with logs that no reviewer can interpret, or claim proprietary opacity. The test is substantive contribution and meaningful evaluability, not labels or data volume. If not evaluable, reliance must be limited or rejected. |
| 12. Boundary review | Does not declare a system high risk under AI-02 or satisfy all AI governance, privacy, security or intellectual-property duties. Those remain separate. It prevents automation from becoming unacknowledged professional authority. |
| 13. Implementation-neutrality review | No required model, vendor, prompt log, agent architecture, MCP/A2A protocol, canonical case file or explainability method. |
| 14. Terminology review | Use delegated or technology-enabled analysis, material contribution, meaningful human review and traceability. Avoid defining AI broadly when the control applies to any tool. |
| 15. Drafting | Proposed provision: Where delegated persons or technology materially contribute to analysis, evidence selection, judgment or drafting of the conclusion, the practitioner shall understand and document their role, material inputs, source provenance, assumptions and known limitations; evaluate and, where necessary, corroborate material outputs; and retain the ability and authority to question, correct or reject them. Human review and output authorisation shall be substantive relative to the contribution and shall not be replaced by automated approval or uncritical reliance. Material limitations that affect interpretation shall be included in the conclusion boundaries. Applicable law and professional requirements shall also be met. |
| 16. Open issue record | OI-20: proprietary opacity; OI-38: disclosure threshold for tool use; OI-39: evidencing meaningful review; OI-40: rapidly changing law and standards. |
DOC-01 — Traceable engagement documentation#
Disposition: requirement; proceed.
Evidence strength: strong by professional analogy.
| Development stage | Record |
|---|---|
| 1. Problem definition | Without a contemporaneous record, the practitioner cannot show what question, information, challenge, judgment, dissent and authorisation actually supported the report. Later narratives are vulnerable to hindsight reconstruction, lost versions and selective justification. |
| 2. Scope relevance | Necessary for professional defensibility and review of the engagement; not a general records-management standard. |
| 3. Source classification | IA-01 Standards 14.6 and related documentation provisions; ASR-08; ASR-01 and ASR-05–06; AI-02–04; DS-04; DD-02–03. |
| 4. Necessity test | Necessary. Most other requirements cannot be observed or independently reviewed without documentation. |
| 5. Non-duplication test | Documentation is well covered in audit and internal audit. Their retention periods, completion rules and experienced-auditor formulations are not portable. This candidate adopts a narrower informed-reviewer understanding test. |
| 6. Normative classification | Requirement. Presence, integrity and comprehensibility of a record are observable. |
| 7. Independence test | General professional condition, independent of any canonical case file. |
| 8. Generalisability test | Paper, document repository, case-management system or other controlled form may be used. It accommodates lawful confidentiality and privilege restrictions. |
| 9. Observability test | Dated/versioned records of qualification, question/scope changes, independence/competence, sources, material work, challenge, judgments, reliance, assumptions, dissent, conclusion development and authorisation; controlled final report. |
| 10. Proportionality test | Detail scales with complexity and risk. The record need not allow mechanical re-performance or identical judgment, but must permit understanding of the basis and significant decisions. |
| 11. Adversarial review | Volume can substitute for traceability; auto-generated logs can be unreadable; records may be created after the result. The test is a coherent contemporaneous reasoning trail, not data accumulation. |
| 12. Boundary review | The Standard does not set a legal retention period, override privilege/confidentiality, create disclosure rights or claim audit documentation. Applicable obligations prevail. |
| 13. Implementation-neutrality review | No repository, file structure, software, immutable ledger or record format prescribed. |
| 14. Terminology review | Use informed independent practitioner rather than experienced auditor. Use understand, not reperform to the same conclusion. |
| 15. Drafting | Proposed provision: The practitioner shall prepare and preserve a controlled engagement record sufficient for an informed independent practitioner, with relevant competence and no prior connection to the engagement, to understand: qualification and acceptance; the decision question, mandate and changes; competence and independence judgments; material sources and provenance; work performed and challenge applied; material assumptions, reliance, counterevidence and dissent; significant professional judgments; the basis and boundaries of the conclusion; and review and output authorisation. Documentation shall be prepared on a timely basis, distinguish versions and identify the authorised report. Retention, access, confidentiality and security shall comply with applicable obligations and be proportionate to the engagement's risk and intended use. |
| 16. Open issue record | OI-41: minimum retention period; OI-42: documentation access for external review; OI-43: privilege/confidentiality; OI-44: dynamic or continuously updated reports. |
PROP-01 — Proportionality with an essential-condition floor#
Disposition: principle, with operative references in requirements; proceed.
Evidence strength: strong.
| Development stage | Record |
|---|---|
| 1. Problem definition | A one-size process can make the Standard bureaucratic and unusable, while an undefined appeal to proportionality can excuse superficial evidence, token challenge or absent safeguards. |
| 2. Scope relevance | Governs the extent and form of all engagement work; it does not define enterprise materiality or risk appetite. |
| 3. Source classification | IA-01 Standard 4.2; RISK-01; DD-01; ASR-01; ETH-01; PRACT-01–04. Mature professional and risk-based practice. |
| 4. Necessity test | Necessary as an architectural principle. Without it, requirements overreach or become easy to evade. |
| 5. Non-duplication test | Proportionality is established elsewhere; the new contribution is a category-specific factor set and explicit floor. No proprietary risk score is created. |
| 6. Normative classification | Principle, because proportionality is a cross-cutting judgment. Specific requirements make its consequences observable. |
| 7. Independence test | Fully generic and externally grounded. |
| 8. Generalisability test | Applies to different sectors, sizes, internal/external models and qualitative/quantitative questions. |
| 9. Observability test | Planning rationale connects effort to materiality, complexity, uncertainty, reversibility, reliance, novelty, timing, independence threats and intended use; report boundaries correspond to residual constraints. |
| 10. Proportionality test | Self-applying. It explicitly cannot remove category existence, adequate competence, acceptable independence, truthful reporting, adequate basis or authority separation. |
| 11. Adversarial review | A provider may label a major decision “low complexity” or use fee/time constraints as the only proportionality factor. The documented rationale and conclusion boundaries must make the judgment reviewable. |
| 12. Boundary review | Does not replace applicable legal/professional materiality or risk models and does not require exhaustive work on a major decision. |
| 13. Implementation-neutrality review | No tiering matrix, points score, hours, team size or review level prescribed. |
| 14. Terminology review | Use proportionate with explicit factors; avoid implying “light assurance”. |
| 15. Drafting | Principle: The nature, timing and extent of work should be proportionate to the reasonably foreseeable significance of the decision, difficulty of reversal, complexity, uncertainty, novelty, material reliance, independence threats, available time and intended use. Proportionality cannot remove the need for a qualifying decision and timely use, adequate competence and basis, acceptable independence, genuine challenge, truthful conclusion boundaries, human output authorisation or preservation of client decision authority. |
| 16. Open issue record | OI-45: empirical calibration of factors; OI-46: emergency decisions; OI-47: whether sector annexes should specify higher floors. |
NONC-01 — Inability to conclude, withholding and withdrawal#
Disposition: requirement; proceed and likely integrate with QUAL-01 and CONC-01 in the final Standard.
Evidence strength: strong by analogy; moderate for generic disposition terminology.
| Development stage | Record |
|---|---|
| 1. Problem definition | Commercial and governance pressure can produce a clean opinion even after independence, competence, access, evidence or timing fails. A Standard that describes only successful completion rewards concealment. |
| 2. Scope relevance | Directly necessary to maintain category integrity from acceptance through release. |
| 3. Source classification | ASR-01 (scope limitations, modified conclusions, withdrawal); ETH-01 (decline/discontinue for unaddressed threats); IA-01; ASR-05–07; DD-02–03. |
| 4. Necessity test | Necessary. Visible inability to conclude is a legitimate professional result and prevents unsupported certainty. |
| 5. Non-duplication test | Assurance and audit use specified opinion forms that cannot be transplanted. The candidate states a general decision rule and does not mimic audit qualification categories. |
| 6. Normative classification | Requirement. The practitioner's response to a failed condition is observable. |
| 7. Independence test | General professional safeguard and not tied to Westbridge House workflow. |
| 8. Generalisability test | Allows lawful options appropriate to internal/external and jurisdictional contexts: remedy, legitimate narrowing/reacceptance, explicitly limited conclusion, no conclusion, withholding or withdrawal. |
| 9. Observability test | Issue and impact assessment; remediation attempts; reacceptance/scope-change record; report limitation/no-conclusion statement; withholding/withdrawal record; communication to appropriate parties. |
| 10. Proportionality test | Response severity matches the failed condition. Essential independence or competence failures ordinarily require decline/withdrawal; a bounded information gap may permit a transparently limited conclusion. |
| 11. Adversarial review | A sponsor may repeatedly narrow the question until a favourable conclusion becomes possible, or issue an unauthorised draft after withdrawal. Narrowing must remain decision-relevant and pass QUAL-01; report status and release control must be clear. |
| 12. Boundary review | Does not prescribe audit opinion labels or override law/contract on withdrawal and communication. It should not imply that every limitation is non-conformity if properly bounded. |
| 13. Implementation-neutrality review | No fixed escalation ladder or approval workflow. |
| 14. Terminology review | Prefer explicitly limited or conditional conclusion and inability to conclude; avoid unqualified/qualified audit opinion. Withdrawal is used subject to applicable rights. |
| 15. Drafting | Proposed provision: If a qualification condition, acceptable independence, necessary competence, mandate/access, adequate conclusion basis, truthful communication or meaningful output authorisation is absent or materially impaired, the practitioner shall not issue a more definitive conclusion or conformity claim than the circumstances support. The practitioner shall, as appropriate, obtain remediation, legitimately revise and reaccept the decision question, issue an explicitly limited or conditional conclusion, state an inability to conclude, withhold the professional output, or withdraw. A revision shall not remove a matter material to the actual consequential decision merely to enable a favourable conclusion. Material effects and the disposition shall be documented and communicated to the appropriate engaging and decision authorities, subject to applicable law. |
| 16. Open issue record | OI-46: urgency/time limits; OI-48: withdrawal rights and duties; OI-49: client use of drafts/withdrawn reports; OI-50: terminology for limited conclusions. |
3. Candidates not admitted as v0.1 requirements#
The following candidates complete the same development cycle. Their rejection or downgrade is part of the normative result.
AQT-01 — Mandatory “Accepted Question” construct#
Disposition: delete as drafted; the legitimate problem is addressed by QUEST-01 using generic terminology.
Evidence strength: practice-derived.
| Development stage | Record |
|---|---|
| 1. Problem definition | Work needs a stable, decision-relevant unit and formal acceptance before substantive analysis. |
| 2. Scope relevance | The underlying problem is in scope; the branded construct is not necessarily required. |
| 3. Source classification | WH-01 (Westbridge House practice); IA-01 and ASR-01–02 (external scope/engagement-term analogies). No external source uses the construct as a general category term. |
| 4. Necessity test | A documented decision question is necessary; calling it Accepted Question or using a particular acceptance object is not. |
| 5. Non-duplication test | Existing objective, scope and engagement-term concepts largely cover the control. QUEST-01 adds only decision-specific elements. AQT-01 would create needless terminology. |
| 6. Normative classification | Delete/replace, not a requirement, principle or recommendation. |
| 7. Independence test | The underlying idea survives Westbridge House; the name and implied workflow do not pass independently. |
| 8. Generalisability test | A mandatory proprietary form would not generalise; decision-question documentation does. |
| 9. Observability test | A signed “Accepted Question” could be observed but would prove form, not decision relevance. QUEST-01 has stronger substantive observations. |
| 10. Proportionality test | A single formal object may be excessive for small work and insufficient for complex mandates. |
| 11. Adversarial review | A sponsor can craft an accepted but biased question and treat signature as proof of adequacy. |
| 12. Boundary review | “Acceptance” can be confused with client acceptance of risk or conclusion; separate engagement acceptance is clearer. |
| 13. Implementation-neutrality review | Fails if it requires Westbridge House's canonical object/workflow. |
| 14. Terminology review | Decision question is plainer, more international and category-specific. |
| 15. Drafting | No provision. The necessary content is restated in QUEST-01 and QUAL-01 without the term. |
| 16. Open issue record | No issue on the term. OI-05–07 remain for decision-question content and changes. |
SEP-01 — Universal separation from all implementation services#
Disposition: downgrade to a conditional recommendation/safeguard under IND-01; not a universal requirement.
Evidence strength: limited to moderate for a universal rule; strong that implementation interests can create threats.
| Development stage | Record |
|---|---|
| 1. Problem definition | A provider that designed the proposal or expects implementation revenue may self-review, advocate or soften an adverse conclusion. |
| 2. Scope relevance | Directly relevant to independence, but not every implementation relationship has the same significance. |
| 3. Source classification | ETH-01; IA-01; IA-02; ASR-01; CONS-01; WH-01 practice. Sources support threat evaluation, not a universal ban for this category. |
| 4. Necessity test | An effective response to a material threat is necessary. Complete organisational separation in every case is not shown necessary. |
| 5. Non-duplication test | IND-01 already requires evaluation, elimination/safeguards/withdrawal. A separate universal rule would override the better-developed threat model without evidence. |
| 6. Normative classification | Recommendation/conditional safeguard, not requirement. A conclusion- or outcome-contingent implementation award is already captured as unacceptable under IND-01. |
| 7. Independence test | The concern is independent; Westbridge House's exact “no implementation sale in the same case” practice is not automatically universal. |
| 8. Generalisability test | A total ban can be impractical for small specialist markets, internal functions or integrated professional services. Separate practitioner/personnel can often generalise. |
| 9. Observability test | Prior/current/future services, revenue incentives, proposal authorship, team separation, governance and fee arrangements. |
| 10. Proportionality test | Response depends on materiality of prior work, decision influence, revenue, common leadership and ability to use another practitioner. Some threats remain unacceptable regardless of safeguards. |
| 11. Adversarial review | Legal-entity or team separation can be cosmetic when incentives and leadership are shared. Conversely, a nominal ban can be bypassed through affiliates. Substance must govern. |
| 12. Boundary review | Too strict a rule could exclude legitimate consultancy inputs or imply the assurance independence code. Too weak a rule lets an adviser certify its own proposal. |
| 13. Implementation-neutrality review | A universal firm-structure rule would fail. A functional safeguard remains neutral. |
| 14. Terminology review | Use implementation interest, self-review and advocacy threat; avoid implementation firewall as a defined term. |
| 15. Drafting | Recommendation: Where the practitioner or its network has materially designed or advocated the proposal, or reasonably expects implementation work, the practitioner should use a separate practitioner or demonstrably independent personnel and governance unless a documented assessment shows that threats are reduced to an acceptable level. An implementation award or fee dependent on the second-opinion conclusion or client decision is not acceptable under IND-01. |
| 16. Open issue record | OI-12–14: prior authorship, cooling-off and implementation interests; practice validation needed across firm types. |
REV-01 — Mandatory separate independent quality reviewer#
Disposition: downgrade to recommendation for selected high-risk engagements; not a core requirement.
Evidence strength: moderate in audit/assurance; limited direct evidence for this category.
| Development stage | Record |
|---|---|
| 1. Problem definition | Engagement leaders can miss errors or become committed to their own analysis; a separate reviewer can challenge significant judgments and release readiness. |
| 2. Scope relevance | Relevant to output quality, but a mandatory separate person could make small or specialist engagements infeasible. |
| 3. Source classification | ASR-06–08 and ASR-05 (formal audit/quality standards); IA-01; academic bias evidence DS-01–09. Direct audit rules apply only to specified engagements. |
| 4. Necessity test | Meaningful human output authorisation is necessary. A distinct independent reviewer for every case is not demonstrated as necessary. |
| 5. Non-duplication test | AUTH-01 already ensures accountable review and release control. Importing an engagement-quality-review model would duplicate an audit construct and imply firm infrastructure. |
| 6. Normative classification | Recommendation where risk factors justify separation. |
| 7. Independence test | The concept is independent of Westbridge House reviewer gates; its universality fails. |
| 8. Generalisability test | Selective independent review is feasible in larger/high-risk work; solo and rare-specialist contexts require alternatives such as external consultation. |
| 9. Observability test | Reviewer appointment, competence/independence, scope, significant judgments reviewed, issues raised/disposed and release concurrence where used. |
| 10. Proportionality test | Factors include stakes, complexity, novelty, authoriser involvement, unresolved dissent, independence threats, opaque tool use and conclusion sensitivity. |
| 11. Adversarial review | A rubber-stamp reviewer or colleague lacking time/competence adds ceremony, not quality. Appointment must not diffuse the output authoriser's responsibility. |
| 12. Boundary review | Must not be called an audit engagement quality review unless the applicable framework uses that term; reviewer does not make the client's decision. |
| 13. Implementation-neutrality review | Conditional recommendation leaves reviewer source, timing and method open. |
| 14. Terminology review | Use independent pre-release review descriptively; avoid concurring reviewer as a category term. |
| 15. Drafting | Recommendation: The practitioner should consider an independent pre-release review when the decision significance, complexity, novelty, independence threats, authoriser involvement, material dissent or opacity of delegated analysis creates heightened risk that AUTH-01 alone will not provide sufficient challenge. The reviewer should be competent, sufficiently independent and unable to assume the output authoriser's responsibility. |
| 16. Open issue record | OI-51: thresholds and empirical benefit; OI-31: solo practitioners; OI-37: internal hierarchy. |
TECHNIQUE-01 — Mandatory red team, premortem or independent model#
Disposition: downgrade to guidance examples under CHAL-01 and UNC-01.
Evidence strength: moderate for selected techniques in selected contexts; limited for universal application.
| Development stage | Record |
|---|---|
| 1. Problem definition | Abstract instructions to “be skeptical” may not overcome confirmation, escalation, shared-information or framing effects. Structured methods can make challenge concrete. |
| 2. Scope relevance | Techniques may support in-scope challenge but are means, not category ends. |
| 3. Source classification | DS-01 (consider opposite); DS-10 (prospective hindsight); DS-12 (adversarial collaboration); PRACT-04 (independent estimates/sensitivity); professional practice. |
| 4. Necessity test | Genuine challenge is necessary. No one named technique is necessary across all subject matters. |
| 5. Non-duplication test | CHAL-01 and UNC-01 specify the performance outcome. A separate rule would duplicate them and privilege a method without comparative evidence. |
| 6. Normative classification | Guidance, not requirement or universal recommendation. |
| 7. Independence test | Techniques exist independently, but a Westbridge House red-team/agent configuration would not. |
| 8. Generalisability test | Applicability varies: premortem may suit project risks, independent model may suit valuations, adversarial collaboration may suit disputed hypotheses. |
| 9. Observability test | Where selected, method, participants, inputs, findings and disposition can be recorded. Completion alone does not show genuine challenge. |
| 10. Proportionality test | Choose based on the material failure mode, data, complexity, time and likely information value. |
| 11. Adversarial review | A ritual red team can be staffed by agreeable insiders; a premortem can generate generic risks; an independent model can repeat the same assumptions. Outcome evidence matters. |
| 12. Boundary review | Techniques do not provide assurance or transfer decision ownership; quantitative models require competence and transparent assumptions. |
| 13. Implementation-neutrality review | Guidance lists alternatives and does not prescribe people, agents, software or templates. |
| 14. Terminology review | Use standard names only as examples; explain function rather than brand. |
| 15. Drafting | Guidance: Depending on the material failure mode, challenge methods may include considering the opposite, independent source search, alternative framing, reference-class comparison, sensitivity/scenario analysis, prospective hindsight (premortem), independent estimation, adversarial review or a separately constituted red team. Selection and quality matter more than completion of a named exercise. |
| 16. Open issue record | OI-25 and OI-27: validation of challenge sufficiency and gaming resistance. |
CONF-01 — Mandatory numerical confidence score#
Disposition: delete; retain as hypothesis/open issue only.
Evidence strength: limited/hypothesis.
| Development stage | Record |
|---|---|
| 1. Problem definition | Users need to understand how strongly the conclusion is supported and where uncertainty lies. A score appears to offer comparability. |
| 2. Scope relevance | Communication of conclusion strength is relevant; numerical scoring is only one possible method. |
| 3. Source classification | DS-09 (calibration evidence); RISK-01–02; PRACT-04; WH-01 practice terminology. No source validates a universal confidence scale for heterogeneous one-off decisions. |
| 4. Necessity test | Explicit boundaries are necessary; a numerical score is not and may mislead. |
| 5. Non-duplication test | Assurance uses defined engagement levels, statistics uses defined intervals, and risk practice may use calibrated probabilities. A generic score would borrow their authority without their methods. |
| 6. Normative classification | Delete as requirement; hypothesis for later research. |
| 7. Independence test | A branded or internally calibrated score would depend on Westbridge House unless independently validated. |
| 8. Generalisability test | Fails across qualitative legal, strategic, technical and market judgments with different uncertainty structures. |
| 9. Observability test | A number is observable but its validity, calibration and comparability are not established. |
| 10. Proportionality test | Apparent simplicity hides modelling burden and false precision. |
| 11. Adversarial review | Scores can be negotiated, anchored, averaged across incomparable dimensions or used as pseudo-assurance. |
| 12. Boundary review | High risk of confusion with assurance level, probability of success, statistical confidence or credit rating. |
| 13. Implementation-neutrality review | A required scale would prescribe a methodology and likely advantage its owner. |
| 14. Terminology review | Use conclusion boundaries and plain descriptions; do not use generic confidence level. |
| 15. Drafting | No provision. CONC-01 requires qualitative or appropriately supported quantitative boundaries without a universal score. |
| 16. Open issue record | OI-52: whether any validated, decision-type-specific calibration method can support future guidance. |
DISS-02 — Mandatory unanimous resolution of all material dissent#
Disposition: delete; DISS-01 requires reasoned disposition and, where necessary, disclosure or withholding.
Evidence strength: limited and conflicting for this category.
| Development stage | Record |
|---|---|
| 1. Problem definition | Contradictory professional views can confuse users or show that the report is not release-ready. Audit quality systems commonly require differences to be resolved. |
| 2. Scope relevance | Material disagreement is in scope; compulsory unanimity is a proposed response. |
| 3. Source classification | ASR-05–07 (formal audit rules); IA-01 (communication of both positions can occur); GOV-02–03; DS-07 and DS-12. Sources support attention, not universal unanimity in non-assurance work. |
| 4. Necessity test | A defensible disposition is necessary. Unanimity is not: genuine uncertainty can leave reasonable professionals divided. |
| 5. Non-duplication test | Importing audit release rules would misapply a specific quality framework. DISS-01 addresses the category gap without forcing an artificial consensus. |
| 6. Normative classification | Delete. It should not be a recommendation either. Resolution is one available disposition. |
| 7. Independence test | The issue is generic; a universal consensus rule does not survive methodological scrutiny. |
| 8. Generalisability test | Fails in multidisciplinary or novel questions where legitimate interpretive pluralism remains. |
| 9. Observability test | A unanimous signature is observable but can result from hierarchy or suppression rather than agreement. |
| 10. Proportionality test | The rule would let one strategic objection block all work, regardless of materiality or available disclosure. |
| 11. Adversarial review | Leaders can coerce withdrawal of dissent; dissenters can exercise a veto; both harm independence. |
| 12. Boundary review | Could falsely emulate audit quality management and imply the final conclusion is objectively certain. |
| 13. Implementation-neutrality review | Would prescribe team governance and voting/consensus expectations. |
| 14. Terminology review | Use resolved, unresolved and disclosed, conclusion modified, withheld or withdrawn as dispositions, not consensus achieved. |
| 15. Drafting | No provision. DISS-01 governs material dissent. |
| 16. Open issue record | OI-28–31 remain; no further issue on unanimity. |
REPERF-01 — Reperformance of all third-party work#
Disposition: delete; replace with risk-based evaluation in REL-01.
Evidence strength: weak as a universal rule and inconsistent with established expert-reliance models.
| Development stage | Record |
|---|---|
| 1. Problem definition | Blind reliance on another adviser, expert or report can import errors, scope gaps and conflicts. Reperformance appears to eliminate the risk. |
| 2. Scope relevance | Reliance is in scope; universal reperformance is a proposed method. |
| 3. Source classification | ASR-04; ETH-02; DD-02–03; IA-01. Mature sources require evaluation and additional procedures, not wholesale reperformance. |
| 4. Necessity test | Evaluating fitness and doing additional work where needed are necessary. Reperforming all specialist work is neither feasible nor necessary. |
| 5. Non-duplication test | REL-01 directly adopts the established risk-based approach; REPERF-01 would contradict it. |
| 6. Normative classification | Delete. Reperformance may be selected as a procedure under REL-01. |
| 7. Independence test | The idea is not Westbridge House-specific, but it lacks independent professional support. |
| 8. Generalisability test | Fails where regulated or rare expertise is required and would undermine legitimate multidisciplinary reliance. |
| 9. Observability test | Reperformance is observable but may reproduce the same inputs/method and still not validate the work. |
| 10. Proportionality test | Grossly disproportionate for immaterial or well-supported work; potentially justified for a dispositive, doubtful calculation. |
| 11. Adversarial review | Superficial “reperformance” can be claimed without independent assumptions; cost can be used to discourage any second opinion. |
| 12. Boundary review | Could cause the practitioner to practise outside competence or imply an audit of all specialist work. |
| 13. Implementation-neutrality review | Overprescribes a procedure. |
| 14. Terminology review | Use evaluate, corroborate, test and perform additional work where necessary. |
| 15. Drafting | No provision. REL-01 allows reperformance when proportionate and necessary. |
| 16. Open issue record | OI-19–21 cover depth of reliance evaluation. |
ARCH-01 — Prescribed canonical case file, agents or technology stack#
Disposition: delete entirely from the public Standard.
Evidence strength: Westbridge House practice only for the specified architecture.
| Development stage | Record |
|---|---|
| 1. Problem definition | Complex engagements need controlled information, provenance, division of work, review, versioning and secure collaboration. A standard architecture can operationalise these needs. |
| 2. Scope relevance | The required outcomes—traceability, controlled delegation and review—are in scope. The technical architecture is not constitutive. |
| 3. Source classification | WH-01 (practice); AI-03–04 and professional documentation sources support outcomes but not canonical case files, agents, MCP, A2A, Decision Signals or a Decision Room. |
| 4. Necessity test | Outcomes are necessary; the specified implementation is not. |
| 5. Non-duplication test | DOC-01, TECH-01 and AUTH-01 already define the observable performance. A technical rule would duplicate them as design prescription. |
| 6. Normative classification | Delete. At most, implementations could appear in separate non-normative case studies, outside v0.1. |
| 7. Independence test | Fails: the specified elements are tied to Westbridge House architecture and would create a hidden vendor advantage. |
| 8. Generalisability test | Fails for paper-based, small-team, internal, low-technology and competing implementations. |
| 9. Observability test | System components are observable, but their existence does not prove good evidence, challenge or human review. |
| 10. Proportionality test | A uniform stack would be excessive for many engagements and could still be inadequate for high-risk work. |
| 11. Adversarial review | A provider can install compliant-looking tools while feeding them biased data and rubber-stamping outputs. |
| 12. Boundary review | Would expand the Standard into software and AI governance, contrary to scope. |
| 13. Implementation-neutrality review | Explicitly fails. |
| 14. Terminology review | Do not include product/workflow names. Use functional terms only where necessary. |
| 15. Drafting | No provision. TECH-01 and DOC-01 are the complete public treatment in v0.1. |
| 16. Open issue record | None for v0.1. Future implementation guidance must remain vendor-neutral and is not a current project. |
CERT-01 — Certification or provider-level conformity#
Disposition: out of scope and delete. Engagement-level conformity wording remains an open issue, not a certification scheme.
Evidence strength: hypothesis.
| Development stage | Record |
|---|---|
| 1. Problem definition | Users may want an easy signal that a provider or engagement meets the Standard. Certification appears to increase trust. |
| 2. Scope relevance | Certification, accreditation, licensing and conformity assessment are expressly outside the project boundary. |
| 3. Source classification | DRAFT-01 provides drafting conventions only; no conformity-assessment source was researched because the subject is out of scope. WH-02 prohibits starting such work. |
| 4. Necessity test | Not necessary to define the category. Premature certification could institutionalise unvalidated requirements. |
| 5. Non-duplication test | Conformity assessment is a separate professional field. Inventing it here would duplicate that field without research. |
| 6. Normative classification | Out of scope/delete. |
| 7. Independence test | A Westbridge House-controlled mark would fail the public Standard's independence premise absent separate governance; not investigated. |
| 8. Generalisability test | Unknown and unvalidated. |
| 9. Observability test | A certificate is observable but says nothing about assessment competence, sampling, surveillance or independence without an entire scheme. |
| 10. Proportionality test | Scheme burden and risk are unknown. |
| 11. Adversarial review | Highest risk of pseudo-certification, pay-to-play approval and marketing capture. |
| 12. Boundary review | Would create reliance, liability, accreditation and competition issues beyond the Standard. |
| 13. Implementation-neutrality review | A provider-controlled scheme could create structural advantage. |
| 14. Terminology review | v0.1 must not use certified, accredited, approved provider or assured compliant. |
| 15. Drafting | No provision. The Standard may only define a carefully limited engagement-level conformity statement if OI-01 is resolved; v0.1 should not authorise a mark or provider certification. |
| 16. Open issue record | OI-01: engagement-level conformity wording. Any future certification/accreditation question requires a separate project and governance decision. |
4. Pre-adversarial admission result#
The development cycle admits sixteen candidates as requirements, one as a conditional requirement, two as principles, two as recommendations and one group of challenge methods as guidance. Nine proposed rules or implementations are not admitted as requirements. The final Standard may merge related accepted candidates, so this is not a target clause count.
The candidates with the weakest requirement-level evidence are DISS-01 and TECH-01. They proceed because each addresses a material and observable category failure that is not adequately addressed by merely referring to another framework. Their thresholds and language require particular attention in the adversarial review.