← All modules
1. Overview2. Risk Register3. 5×5 Matrix4. Governance Triangle5. Assessment History6. Ownership7. Before vs After8. Receipts
TOURWB-MOD-006 · Architecture ratified 2026-04-16 · Tour shipped 2026-04-16

Risk Register

Controls define what should work. Findings document what didn't. Risk quantifies why it matters.

Risk Register started as a dimensional table inside AuditForge, tracking likelihood and impact as mutable fields that silently overwrote history. When the concept proved universal, it was extracted to WorkBench. But unlike Controls and Findings, Risk didn't just move. It matured. The promotion split one table into seven fact-disciplined artifacts. The data migrated. The overwrite semantics didn't.

Risks
17
14 extracted · 3 extended
Assessments
20
Immutable fact stream
Unmitigated
10
Zero controls mapped
Categories
8
All 8 represented
Triangle
Complete
Risk → Controls → Findings
Governance Triangle Complete
Three promoted modules, three cross-promoted sibling relationships. Controls · Findings · Risk. One substrate. No integration layer.

A risk register is not a list of what might go wrong. It is a list of what you are watching so you can act early.

All modulesThe 5×5 matrix ↓
Section 2

The Risk Register

17 risks across 8 categories. 10 have no controls mapped.

Try this: scan for the “Unmitigated” badges — those are risks with no controls yet.
RISK-001Extracted
CRITICALASSESSED
Critical context lost or corrupted during long conversations due to compaction, token limits, or platform interruption.
CybersecurityScore: 20Controls: 1Owner: CTO
RISK-002ExtractedUnmitigated
MEDIUMASSESSED
Context from one project contaminates another causing misattribution, scope confusion, or data leakage between engagements.
OperationalScore: 9Controls: 0Owner: Priya Shankar EMP-003
RISK-003ExtractedUnmitigated
HIGHASSESSED
The DDL Audit Universe framework is not independently reviewed, allowing structural errors, drift, or bias to persist undetected.
OperationalScore: 12Controls: 0Owner: Maya Okafor EMP-001
RISK-004ExtractedUnmitigated
HIGHACCEPTED
DDL operations depend on a single individual. Incapacitation, departure, or unavailability creates existential risk.
StrategicScore: 15Controls: 0Owner: CSO
RISK-005Extracted
MEDIUMMITIGATED
Control testing evidence not retained with defined standards, rendering controls unprovable under external audit.
ComplianceScore: 9 → 6Controls: 2Owner: Clara Nilsson EMP-005
RISK-006ExtractedUnmitigated
HIGHASSESSED
AI-generated artifact treated as canonical without human review. Builder-confirms-all boundary degrades over time.
ComplianceScore: 15Controls: 0Owner: Maya Okafor EMP-001
RISK-007Extracted
HIGHMONITORING
Human review threshold informally rises as operator becomes comfortable with AI outputs. Boundary degradation occurs gradually.
OperationalScore: 12 → 8Controls: 3Owner: Jordan Webb EMP-004
RISK-008Extracted
HIGHASSESSED
Prompts deployed to production without version control, review, or approval. Prompt drift causes inconsistent outputs.
OperationalScore: 15Controls: 1Owner: CTO
RISK-009ExtractedUnmitigated
HIGHASSESSED
Prompt produces outputs that do not match intended behavior due to ambiguous instructions or missing constraints.
OperationalScore: 12Controls: 0Owner: Priya Shankar EMP-003
RISK-010Extracted
HIGHASSESSED
Sensitive client data inadvertently included in prompts sent to external AI providers. Data exposure risk.
ComplianceScore: 10Controls: 2Owner: Jordan Webb EMP-004
RISK-011ExtractedUnmitigated
HIGHASSESSED
AI model outputs used to make financial, legal, or strategic decisions without human expert review.
OperationalScore: 12Controls: 0Owner: Hana Takeda EMP-007
RISK-012ExtractedUnmitigated
MEDIUMIDENTIFIED
Control framework grows beyond operational capacity, creating compliance overhead that diverts resources from core business.
FinancialScore: 6Controls: 0Owner: CFO
RISK-013Extracted
CRITICALASSESSED
Unauthorized access to financial systems due to stale user accounts and incomplete deprovisioning processes.
CybersecurityScore: 20 → 12Controls: 2Owner: Priya Shankar EMP-003
RISK-014Extracted
MEDIUMASSESSED
DDL deliverables incorporate material AI-generated content without appropriate disclosure, violating client contracts or regulations.
ComplianceScore: 6Controls: 1Owner: Theo Grant EMP-009
RISK-015Tour ExtensionUnmitigated
MEDIUMASSESSED
Negative media coverage or social media incident damages DDL brand, causing client attrition or talent loss.
ReputationalScore: 5Controls: 0Owner: CSO
RISK-016Tour ExtensionUnmitigated
MEDIUMASSESSED
Remote work ergonomic issues lead to repetitive strain injuries among extended-hours engineering staff.
Health & SafetyScore: 5Controls: 0Owner: Theo Grant EMP-009
RISK-017Tour ExtensionUnmitigated
MEDIUMASSESSED
Cloud infrastructure energy consumption exceeds sustainability commitments as compute scales with AI workloads.
EnvironmentalScore: 6Controls: 0Owner: CTO
Section 3 — Featured

The 5×5 Risk Matrix

Every risk plotted by likelihood × impact. Color + numeric score in every cell. Three visual states: unmitigated, inherent+residual, controls-mapped-awaiting-assessment.

Try this: find the risk in the top-right corner. That's the one that keeps the CFO up at night.
RareUnlikelyPossibleLikelyAlmost Certain
Severe
5
10
15
20
25
Major
4
8
12
16
20
Moderate
3
6
9
12
15
Minor
2
4
6
8
10
Negligible
1
2
3
4
5
● Solid = unmitigated or residual assessed◌ Dashed = controls mapped, awaiting residual
Section 4 — Featured

The Governance Triangle

Risk drives controls. Controls are tested. Findings MAY trigger reassessment. Three modules, one loop.

Try this: follow a risk through the entire circuit — from assessment to control to finding to reassessment.
Guided Exemplar · RISK-013 → CTRL-003 → FND-001 → Reassessment
1
Risk Assessed
RISK-013: Unauthorized access to financial systems due to stale user accounts and incomple…
9/1/2025: Likely × Major = 16 (High). Residual: 8
2
Controls Mapped
CTRL-003 (User Access Provisioning) + CTRL-001 (Privileged Access Review)
Bridge effective from 6/1/2025 — controls existed before assessment
3
Finding Discovered
FND-001: Stale user accounts in financial reporting system
1/10/2026: Medium severity. 14 stale accounts found.
4
Risk Reassessed
INCIDENT_TRIGGERED reassessment after finding
3/1/2026: Likely × Severe = 20 (Critical). Impact upgraded to SEVERE.
5
Circuit Closes
Risk drives controls. Controls are tested. Findings update risk.
Three modules, one loop. The Governance Triangle is complete.
Referential Integrity
The substrate protects relationships. Mapped controls and active assessments are preserved. Bridge_RiskControl references are maintained. The triangle holds because the FKs hold.
Section 5

Assessment History

Each assessment is a new row. The old row is unchanged. No overwrite. This is what the new model enables.

Try this: compare a migrated assessment to a native one. The migrated date is backfilled. The native date is real.
Assessment Timeline · RISK-013
INITIALBridgeHIGH
Likely × Major = 16 → residual 8
Initial assessment. Access provisioning controls (CTRL-003, CTRL-001) mapped. Residual assessed at UNLIKELY×MAJOR=8.
9/1/2025 · Priya Shankar EMP-003
INCIDENT_TRIGGEREDNativeCRITICAL
Likely × Severe = 20 → residual 12
Finding FND-001 (stale user accounts, 14 affected) triggered reassessment. Likelihood stays LIKELY. Impact upgraded to SEVERE: financial system access confirmed. Residual reassessed at POSSIBLE×MAJOR=12.
3/1/2026 · Jordan Webb EMP-004 · supersedes FRA-013a
Migrated assessments are marked BACKFILLED_SINGLE. They represent the state as of migration, not the full history. New assessments are fully native.
Section 6

Ownership & Coverage

Six modules, one identity. The tab count IS the Sibling Mandate proof.

Try this: click Maya's name to see her across six modules.
Priya Shankar
Senior Engineer
3
Maya Okafor
Operator
2
Jordan Webb
Engineer
2
Theo Grant
Operations Manager
2
Clara Nilsson
Senior Auditor
1
Hana Takeda
Junior Auditor
1
Control Coverage
10
Unmitigated
6
Partially mitigated
1
Well-mitigated
Independence Proof
RISK-013 is owned by Priya Shankar EMP-003. The reassessment was performed by Jordan Webb EMP-004. Different people, same substrate.
Section 7

Before vs After

Same risk. Better architecture. The promotion didn't just move data — it matured the model.

Try this: count how many rows exist for one risk. In the old model, there was always one. In the new model, there's one per assessment.
Before Promotion
Risk IDRISK-005
DescriptionControl testing evidence not retained with defined standards…
CategoryCOMPLIANCE (enum)
LikelihoodLIKELY (mutable field)
ImpactMODERATE (mutable field)
RatingHIGH (mutable)
Last Updated2026-03-15 (overwrite — old value gone)
1 row. History lost on every update.
After Promotion
Risk IDRISK-005
DescriptionControl testing evidence not retained with defined standards…
Category FKCompliance
Assessments (2 immutable facts):
#1 6/1/2025 · Likely×Moderate=12 · INITIAL
#2 3/15/2026 · Possible×Moderate=9 · PERIODIC (supersedes #1)
2 rows. Full history preserved.
Translation Map
Before→After
likelihood (mutable field)→Fact_RiskAssessment.likelihoodId (FK)
impact (mutable field)→Fact_RiskAssessment.impactId (FK)
inherentRiskRating (mutable)→inherentRiskScore (derived, immutable)
updatedAt (overwrite timestamp)→effectiveDate + recordedDate (bitemporal)
(no history)→previousAssessmentId (supersede chain)

Same risk. Better architecture.

Section 8

Build Receipts

Council reviews, extraction history, the Governance Triangle proof.

CR-WB-RISK-001Ratified 2026-04-16
Risk Register — Architecture
Extraction ratification

Maturation split: dim_risk (1 table, overwrite) → 7 fact-disciplined artifacts. 5 migrations, 18/18 verify.

CR-WB-RISK-TOUR-001Ratified 2026-04-16
Risk Register — Tour
Tour specification

8/8 council convergence. 5×5 matrix, Governance Triangle circuit, maturation before/after, temporal coherence.

Invariant Checklist
✓R1riskId unique per company
✓R2Owner mutex (at most one)
✓R3Every risk has valid categoryId
✓R7Every risk has ≥1 status change
✓R9Every assessment has likelihood + impact
✓R11Score = likelihood.rank × impact.rank
✓SVSeverity correctly mapped from score
✓BTBitemporal fields on all artifacts
Extraction & Maturation Receipt
Migrations 1-4Dim_RiskCategory, Dim_Likelihood, Dim_Impact, Dim_Risk restructure, Fact_RiskAssessment + backfill, Fact_RiskStatusChange + seed, Bridge_RiskControl (6f38059)
Migration 5App code + verification (45a6f64)
Verification18/18 PASS
Maturationdim_risk (1 table, overwrite) → 7 fact-disciplined artifacts. Zero fields use overwrite semantics.
Governance Triangle Receipt
Three promoted modules, three cross-promoted sibling relationships:
Risk → Controls (Bridge_RiskControl) · Controls → Findings (Fact_Finding.controlId) · Findings → Risk (reassessment trigger).
One substrate. No integration layer.
Factlayer → Module Charter → Module Build → HR & People → Payroll & Comp → Time & Attendance → Controls → FindingsRisk Register (you are here)

Third module promoted. First maturation split shipped. Governance Triangle complete. The cathedral compounds.

Sibling · WB-MOD-004
Controls & Compliance →
Sibling · WB-MOD-005
Findings & Observations →