01 — Quick Reference
Menu map
| Group | Models |
|---|---|
| AI Register | Entity · AISystem · AIUseCase · DataSource |
| Assessment & Ethics | RiskAssessment · EthicalRisk · EthicalPrincipleAssessment · LegalComplianceAssessment |
| Controls & Approval | Control · HumanOversightControl · TransparencyObligation · ApprovalGate |
| Evidence & Monitoring | EvidenceRequirement · EvidenceSubmission · MonitoringMetric · MonitoringObservation |
| Incidents, Change & Retirement | AdverseOutcomeIncident · FeedbackComplaint · AIChange · TrainingRequirement · TrainingCompletion · RetirementPlan |
Every model (22) — one line each
| Model | It records | Key foreign keys |
|---|---|---|
| Entity | An audited organisation | — |
| AISystem | An AI system in scope | Entity |
| AIUseCase | A use of the system | AISystem |
| DataSource | Data feeding the system | AISystem |
| RiskAssessment | A lifecycle-stage risk assessment | AISystem, AIUseCase, (self: supersedes) |
| EthicalRisk | An ethical risk | AISystem, RiskAssessment |
| EthicalPrincipleAssessment | A rating against an ethics principle | RiskAssessment |
| Control | A control mitigating a risk | AISystem, EthicalRisk |
| HumanOversightControl | A human intervention point | AISystem, Control |
| TransparencyObligation | A disclosure obligation | AISystem |
| LegalComplianceAssessment | A legal/privacy compliance check | AISystem |
| EvidenceRequirement | Required evidence | AISystem, RiskAssessment, EthicalRisk, Control |
| EvidenceSubmission | Evidence provided | AISystem, EvidenceRequirement, (self: supersedes) |
| ApprovalGate | A lifecycle approval decision | AISystem |
| MonitoringMetric | A metric with thresholds | AISystem |
| MonitoringObservation | An observed value | MonitoringMetric |
| AdverseOutcomeIncident | A harm / near-miss | AISystem |
| FeedbackComplaint | Public feedback/complaint | AISystem, AdverseOutcomeIncident |
| AIChange | A change to the system | AISystem |
| TrainingRequirement | Required training | AISystem |
| TrainingCompletion | Completion of training | TrainingRequirement |
| RetirementPlan | Controlled decommission | AISystem |
Status vocabularies
- System lifecycle: DISCOVERY · DESIGN · BUILD · TEST · PILOT · DEPLOYMENT · OPERATION · CHANGE · RETIREMENT
- System status: PROPOSED · ASSESSING · APPROVED · PILOT · DEPLOYED · SUSPENDED · RETIREMENT_PENDING · RETIRED · REJECTED
- Risk tier / severity: LOW · MEDIUM · HIGH · CRITICAL
- Assessment outcome: SUITABLE · CONDITIONALLY_SUITABLE · NOT_SUITABLE · REASSESSMENT_REQUIRED
- Ethical-risk status: OPEN · TREATMENT_IN_PROGRESS · MONITORING · ACCEPTED · RESOLVED · CLOSED
- Control status / effectiveness: PLANNED·ACTIVE·INEFFECTIVE·RETIRED / NOT_ASSESSED·EFFECTIVE·PARTIALLY_EFFECTIVE·INEFFECTIVE·NOT_APPLICABLE
- Transparency status: NOT_ASSESSED · NOT_REQUIRED · REQUIRED · PUBLISHED · INCOMPLETE
- Approval-gate outcome: APPROVED · CONDITIONAL · REJECTED · DEFERRED
- Monitoring status: NOT_STARTED · ACTIVE · WARNING · BREACH · SUSPENDED · CLOSED
- Incident status: OPEN · TRIAGED · INVESTIGATING · CONTAINED · REMEDIATING · RESOLVED · CLOSED
- Complaint status: RECEIVED · ACKNOWLEDGED · INVESTIGATING · ACTION_REQUIRED · RESOLVED · CLOSED
- Change status: PROPOSED · ASSESSMENT_REQUIRED · APPROVED · IMPLEMENTED · REJECTED · CANCELLED
Demo data set (38 rows)
Entity 1 · AISystem 3 · AIUseCase 2 · DataSource 2 · RiskAssessment 2 · EthicalRisk 2 · EthicalPrincipleAssessment 5 · Control 2 · HumanOversightControl 1 · TransparencyObligation 2 · LegalComplianceAssessment 2 · EvidenceRequirement 1 · EvidenceSubmission 1 · ApprovalGate 2 · MonitoringMetric 2 · MonitoringObservation 2 · AdverseOutcomeIncident 1 · FeedbackComplaint 1 · AIChange 1 · TrainingRequirement 1 · TrainingCompletion 1 · RetirementPlan 1.
Reseed from empty: python apps/aigerm/scripts/seed_data.py (skips if data already present).
Local run
python server/run.py aigerm # http://localhost:5077