Weakener catalog (v0.17.0)

Generated from uofa catalog --format md at 2026-08-30T02:53:21Z. 57 patterns across 5 packs.

Pack: core (23 patterns)

Pattern	Severity	Description
COMPOUND-01	Critical	Critical and High severity weakeners coexist — compounding risk escalation.
COMPOUND-03	High	Assurance level is not Low, yet Critical weakeners exist — stated assurance level may be overstated.
W-AL-01	High	Validation result has no uncertainty quantification — aleatory uncertainty is uncharacterized.
W-AL-02	Medium	Uncertainty quantification is reported but no sensitivity analysis is documented — the drivers of uncertainty are uncharacterized.
W-AR-01	Critical	Credibility factor has a required level but no acceptance criteria — basis for the requirement is unsupported.
W-AR-02	Critical	Credibility factor achievedLevel is below requiredLevel, yet the decision outcome is Accepted with no documented offset rationale — contradictory result ignored.
W-AR-03	High	Requirement requires a specific verification method but the supporting activity used a different activityType — evidence does not answer the question the requirement asked.
W-AR-04	High	ModelConfiguration was validated at a modelVersion that differs from the UofA currentModelVersion — validation evidence may not apply to the current model.
W-AR-05	High	Validation result has no comparedAgainst link — comparator data source is absent.
W-CON-01	High	Decision is Accepted but a credibility factor has neither requiredLevel nor achievedLevel — the acceptance rests on an unestablished factor.
W-CON-02	Medium	UofA references an identifier whose target does not resolve within the graph and has no documented external-fetch hint.
W-CON-03	High	Evidence timestamp is later than the UofA signature timestamp — evidence was created after the UofA was signed, either a clock anomaly or an uncounted update.
W-CON-04	Medium	UofA conforms to ProfileComplete but declares no SensitivityAnalysis — a Complete profile is structurally expected to document sensitivity analysis alongside uncertainty quantification.
W-CON-05	High	VerificationActivity is declared on the UofA but no Evidence is linked via prov:wasGeneratedBy — the activity is a procedural placeholder rather than real verification.
W-EP-01	Critical	Claim has no prov:wasDerivedFrom link to evidence — provenance chain is broken.
W-EP-02	High	Validation result has no prov:wasGeneratedBy — generation activity is missing.
W-EP-03	High	Input dataset vintage predates the current model revision date — input data may be stale relative to the model it exercises.
W-EP-04	High	Credibility factor is not assessed but model risk level exceeds 2 — unassessed factors at elevated risk weaken the credibility argument.
W-ON-01	Critical	UofA has no Context of Use — intended use and risk context are undefined.
W-ON-02	High	Context of Use has neither an applicability constraint nor an operating envelope — the COU is declared but its boundary of validity is undocumented.
W-PROV-01	Critical	Provenance chain terminates at a node that has no upstream derivation/generation/use edge and is not marked uofa:isFoundationalEvidence=true — chain is incomplete.
W-SI-01	Medium	UofA has no digital signature — structural integrity cannot be verified.
W-SI-02	High	UofA is missing bindsRequirement — incomplete profile binding.

Pack: iso42001 (15 patterns)

Pattern	Severity	Description
W-AIMS-AUDIT-STALE	High	Audit results record declares cadence but lacks nextReviewDate — staleness relative to system change frequency cannot be verified.
W-AIMS-CROSSWALK-INVALID	Medium	SoA asserts framework crosswalk but no AIMSControl carries the matching mappedToFramework — crosswalk invalid.
W-AIMS-DATA-DRIFT-UNDETECTED	High	Data resource declared but no OngoingDataQualityMonitoring evidence present — data drift would go undetected.
W-AIMS-DATA-LINEAGE-BROKEN	High	Data resource provenance stage has no hasNextStage successor — lineage discontinuity across processing stages.
W-AIMS-DEPLOYMENT-DRIFT	High	Deployment configuration differs from validated configuration — deployment drift.
W-AIMS-IMPACT-SCOPE	High	Impact assessment scope does not match assessed deployment scope — structural subset mismatch.
W-AIMS-IMPACT-STAKEHOLDER	Medium	Impact assessment record lacks affectedStakeholder enumeration — affected parties not documented.
W-AIMS-INCIDENT-UNCLOSED	High	Nonconformity record present but no root cause analysis linked — incident not closed.
W-AIMS-MODEL-EVAL-SCOPE	High	Model evaluation testSetCoverage does not match documented deploymentPopulation — test set scope mismatch.
W-AIMS-MODEL-EVAL-STALE	High	Model evaluation report version does not match currently deployed model version — evaluation is stale.
W-AIMS-OBJECTIVE-UNMEASURED	Medium	AIMS objective declared but no targetMeasure documented — objective is unmeasured.
W-AIMS-ROLE-UNASSIGNED	High	AIMS role assignment lacks both assignedPerson and assignedOrganizationalUnit — role is unassigned.
W-AL-02	Medium	Claim asserts framework mapping that no other bundle artifact references — claim-evidence alignment broken.
W-AR-02	High	Accept-decision risk has acceptance rationale but no justification body — risk acceptance is not documented.
W-PROV-01	High	Data resource provenance record lacks sourceReference — upstream document or system identifier missing.

Pack: model-credibility (10 patterns)

Pattern	Severity	Description
COMPOUND-EV-01	Critical	Score informs a high-risk decision while carrying neither an uncertainty estimate nor a null baseline - there is no basis for judging whether the reported difference is real.
COMPOUND-EV-02	Critical	A generalized performance claim rests on a sample with no stated relationship to the population it generalizes over - the claim is broader than the evidence supporting it.
W-EV-CAP-06	Medium	No capability-confound control stated - the reported separation may be attributable to general capability rather than to the construct the benchmark claims to measure.
W-EV-COR-09	Medium	No independently furnished measurement corroborates this reported score - the furnished evidence available measures different properties, so the published number stands on a single referent.
W-EV-COU-05	Critical	Reported evaluation states no context-of-use relevance, and an assessment context was supplied - the score is being read against a decision the published record never claimed it informs.
W-EV-DET-03	High	Reported evaluation states no determinism floor - sampling settings and per-model run-to-run spread are uncharacterized, so the score is one draw reported as a measurement.
W-EV-DIV-07	High	The score reported in the published record disagrees with an independent measurement of the same benchmark by more than the measurement tolerance - the two do not agree, and which is correct is not established here.
W-EV-GEN-02	High	Reported benchmark score carries no account of how its items sample the target population - benchmark accuracy cannot be read as generalized accuracy.
W-EV-NULL-04	High	Reported score is not calibrated against a null, chance, or comprehension-free baseline - the floor the score must clear to mean anything is unstated.
W-EV-SUB-08	High	The evaluation subject is not configuration-controlled - its identity is claimed by the provider with no immutable version guarantee, so the score is evidence about an occasion rather than about a fixed artifact.

Pack: nasa-7009b (6 patterns)

Pattern	Severity	Description
W-NASA-01	High	Data pedigree factor is not assessed but model risk level exceeds 1 — input data provenance should be documented at elevated risk.
W-NASA-02	High	Technical review factor is assessed but has no linked ReviewActivity evidence.
W-NASA-03	High	Process management factor is assessed but has no linked ProcessAttestation evidence.
W-NASA-04	Medium	Use history is not assessed but model risk level exceeds 2 — prior deployment evidence would strengthen the credibility argument.
W-NASA-05	High	Results uncertainty factor is assessed but UofA has no UQ data (hasUncertaintyQuantification is missing).
W-NASA-06	High	Results robustness factor is assessed but has no linked SensitivityAnalysis evidence.

Pack: surrogate (3 patterns)

Pattern	Severity	Description
W-SURR-01	High	Declared physics constraint has no linked constraint-check evidence — the surrogate claims to respect a governing law with no checkable residual/conservation artifact.
W-SURR-02	Critical	Surrogate inherits credibility from a parent COU whose recorded decision is Not Accepted — inherited-credibility defeater.
W-SURR-03	High	COU evaluation point/region is not contained within the declared training envelope — the surrogate is exercised in extrapolation.