AI Governance & Management (AIG) Service Line Skills

AI Governance & Management (AIG) Service Line Skills

Service Line Code: AIG Description: ISO 42001:2023, AI readiness, governance frameworks, bias testing, policy development Version: 1.0 Last Updated: 2026-02-02

---

Service Line Overview

AI Governance helps organizations implement responsible AI practices through:

• AI Management System (ISO 42001:2023)
• AI Risk Management
• Ethical AI Frameworks
• Model Governance & Documentation
• Regulatory Compliance (EU AI Act, etc.)

---

Available Skills

#	Skill	Command	Purpose
1	AI Readiness Assessment	/aig-readiness	Assess organization's AI readiness
2	ISO 42001 Gap Analysis	/aig-iso42001-gap	Gap analysis against ISO 42001:2023
3	AI Risk Assessment	/aig-risk	Identify and assess AI-specific risks
4	Bias Testing Framework	/aig-bias-test	Design bias testing approach
5	AI Policy Generator	/aig-policy	Generate AI governance policies
6	AI Ethics Framework	/aig-ethics	Develop responsible AI guidelines
7	Model Risk Management	/aig-mrm	Model risk management framework
8	AI Inventory Manager	/aig-inventory	Catalog and classify AI systems

---

Skill 1: AI Readiness Assessment (/aig-readiness)

Purpose

Assess an organization's readiness to adopt, deploy, and govern AI systems across 6 dimensions.

Assessment Dimensions

Dimension	Weight	Description
Strategy & Vision	15%	AI strategy alignment with business goals
Data & Infrastructure	20%	Data quality, availability, and tech infrastructure
Talent & Skills	20%	AI/ML expertise and training programs
Governance & Ethics	20%	Policies, oversight, and ethical frameworks
Processes & Operations	15%	AI development lifecycle and MLOps
Culture & Change	10%	Organizational readiness for AI adoption

Maturity Levels

Level	Score	Description
1 - Initial	0-20%	Ad-hoc AI initiatives, no formal governance
2 - Developing	21-40%	Pilot projects, emerging governance
3 - Defined	41-60%	Documented processes, governance framework
4 - Managed	61-80%	Metrics-driven, consistent governance
5 - Optimizing	81-100%	Continuous improvement, industry leader

Assessment Questions (Sample)

Strategy & Vision (15%)

1. Does the organization have a documented AI strategy? (0-5)
2. Is AI strategy aligned with business objectives? (0-5)
3. Is there executive sponsorship for AI initiatives? (0-5)
4. Are AI investment priorities defined? (0-5)

Data & Infrastructure (20%)

1. Is data quality measured and monitored? (0-5)
2. Is data infrastructure scalable for AI workloads? (0-5)
3. Are data governance policies in place? (0-5)
4. Is there a data catalog or metadata management? (0-5)

Talent & Skills (20%)

1. Does the organization have AI/ML specialists? (0-5)
2. Is there an AI training program for employees? (0-5)
3. Are AI roles and responsibilities defined? (0-5)
4. Is there a plan to address AI skill gaps? (0-5)

Governance & Ethics (20%)

1. Is there an AI governance framework? (0-5)
2. Are AI ethics principles documented? (0-5)
3. Is there an AI ethics committee or review board? (0-5)
4. Are AI risks formally assessed and managed? (0-5)

Processes & Operations (15%)

1. Is there a standard AI development lifecycle? (0-5)
2. Are MLOps practices implemented? (0-5)
3. Is model versioning and documentation required? (0-5)
4. Are AI systems monitored for drift and performance? (0-5)

Culture & Change (10%)

1. Is there organizational support for AI adoption? (0-5)
2. Are employees comfortable working with AI? (0-5)
3. Is there a change management plan for AI? (0-5)

Output Format

═══════════════════════════════════════════════════════════════════════════════
AI READINESS ASSESSMENT REPORT
Organization: [Name]
Assessment Date: [Date]
Assessor: Core Path Strategies
═══════════════════════════════════════════════════════════════════════════════

OVERALL READINESS SCORE: [XX]% - Level [N]: [Level Name]

DIMENSION SCORES:
─────────────────────────────────────────────────────────────────────────────
Strategy & Vision      ████████░░░░░░░░░░░░  40%  Developing
Data & Infrastructure  ██████████████░░░░░░  70%  Managed
Talent & Skills        ██████░░░░░░░░░░░░░░  30%  Developing
Governance & Ethics    ████░░░░░░░░░░░░░░░░  20%  Initial
Processes & Operations ████████████░░░░░░░░  60%  Defined
Culture & Change       ████████░░░░░░░░░░░░  40%  Developing
─────────────────────────────────────────────────────────────────────────────
WEIGHTED AVERAGE:      ████████░░░░░░░░░░░░  43%  Defined

KEY FINDINGS:
• [Finding 1]
• [Finding 2]
• [Finding 3]

PRIORITY RECOMMENDATIONS:
1. [Recommendation 1]
2. [Recommendation 2]
3. [Recommendation 3]

ROADMAP TO LEVEL [N+1]:
[Phased improvement plan]

═══════════════════════════════════════════════════════════════════════════════

---

Skill 2: ISO 42001 Gap Analysis (/aig-iso42001-gap)

Purpose

Assess compliance against ISO/IEC 42001:2023 - AI Management System standard.

ISO 42001:2023 Structure

Clause	Title	Requirements
4	Context of the Organization	4.1-4.4
5	Leadership	5.1-5.3
6	Planning	6.1-6.3
7	Support	7.1-7.5
8	Operation	8.1-8.4
9	Performance Evaluation	9.1-9.3
10	Improvement	10.1-10.2
Annex A	AI Controls (39 controls)	A.2-A.10
Annex B	AI Objectives (9 categories)	B.2-B.10

Annex A Control Categories

Category	Controls	Description
A.2	Policies for AI	AI policy requirements
A.3	Internal Organization	Roles, responsibilities, competence
A.4	Resources for AI Systems	Data, tools, infrastructure
A.5	Assessing AI System Impacts	Impact assessment process
A.6	AI System Lifecycle	Development, deployment, monitoring
A.7	Data for AI Systems	Data management and quality
A.8	Information for Interested Parties	Transparency and communication
A.9	Use of AI Systems	Responsible use policies
A.10	Third-Party Relationships	Vendor and partner management

Gap Assessment Scoring

Score	Status	Description
0	Not Implemented	No evidence of implementation
1	Planned	Documented plan exists
2	Partially Implemented	Some implementation, not complete
3	Implemented	Fully implemented
4	Managed	Implemented and measured
5	Optimized	Continuous improvement

Output Format

═══════════════════════════════════════════════════════════════════════════════
ISO 42001:2023 GAP ANALYSIS REPORT
Organization: [Name]
Assessment Date: [Date]
═══════════════════════════════════════════════════════════════════════════════

OVERALL COMPLIANCE: [XX]%

CLAUSE COMPLIANCE:
─────────────────────────────────────────────────────────────────────────────
Clause 4: Context           ████████████████░░░░  80%  [X] gaps
Clause 5: Leadership        ████████████░░░░░░░░  60%  [X] gaps
Clause 6: Planning          ██████████░░░░░░░░░░  50%  [X] gaps
Clause 7: Support           ████████████████░░░░  80%  [X] gaps
Clause 8: Operation         ████████░░░░░░░░░░░░  40%  [X] gaps
Clause 9: Performance       ██████░░░░░░░░░░░░░░  30%  [X] gaps
Clause 10: Improvement      ████████░░░░░░░░░░░░  40%  [X] gaps
─────────────────────────────────────────────────────────────────────────────

ANNEX A CONTROLS:
─────────────────────────────────────────────────────────────────────────────
A.2 Policies               ████████████░░░░░░░░  60%
A.3 Organization           ██████████████░░░░░░  70%
A.4 Resources              ████████████████░░░░  80%
A.5 Impact Assessment      ██████░░░░░░░░░░░░░░  30%
A.6 Lifecycle              ████████░░░░░░░░░░░░  40%
A.7 Data                   ████████████░░░░░░░░  60%
A.8 Information            ████░░░░░░░░░░░░░░░░  20%
A.9 Use                    ██████████░░░░░░░░░░  50%
A.10 Third-Party           ██████░░░░░░░░░░░░░░  30%
─────────────────────────────────────────────────────────────────────────────

CRITICAL GAPS (Must Address):
1. [Gap 1] - Clause X.X
2. [Gap 2] - Clause X.X
3. [Gap 3] - Control A.X.X

MAJOR GAPS:
[List of major gaps]

MINOR GAPS:
[List of minor gaps]

REMEDIATION ROADMAP:
─────────────────────────────────────────────────────────────────────────────
Phase 1 (0-3 months): [Critical gap closure]
Phase 2 (3-6 months): [Major gap closure]
Phase 3 (6-9 months): [Minor gaps and optimization]
Phase 4 (9-12 months): [Certification readiness]

ESTIMATED EFFORT: [X] person-months
ESTIMATED COST: $[X]

═══════════════════════════════════════════════════════════════════════════════

---

Skill 3: AI Risk Assessment (/aig-risk)

Purpose

Identify, assess, and prioritize AI-specific risks using NIST AI RMF and ISO 42001 frameworks.

AI Risk Categories

Category	Description	Examples
Technical	Algorithm and model risks	Bias, drift, accuracy, robustness
Data	Data-related risks	Quality, privacy, security, bias
Operational	Deployment and use risks	Availability, performance, misuse
Legal/Compliance	Regulatory risks	GDPR, EU AI Act, sector-specific
Ethical	Societal impact risks	Fairness, transparency, accountability
Strategic	Business risks	Reputation, competitive, financial

Risk Assessment Matrix

Probability	Impact: Negligible	Minor	Moderate	Major	Severe
Almost Certain	Medium	High	High	Critical	Critical
Likely	Low	Medium	High	High	Critical
Possible	Low	Medium	Medium	High	High
Unlikely	Low	Low	Medium	Medium	High
Rare	Low	Low	Low	Medium	Medium

AI-Specific Risk Factors

Bias & Fairness Risks

• Training data bias
• Algorithmic bias
• Proxy discrimination
• Disparate impact

Transparency Risks

• Black-box models
• Unexplainable decisions
• Lack of audit trail

Security Risks

• Adversarial attacks
• Data poisoning
• Model theft
• Prompt injection

Privacy Risks

• Personal data in training
• Inference attacks
• Data leakage

Output Format

═══════════════════════════════════════════════════════════════════════════════
AI RISK ASSESSMENT REPORT
Organization: [Name]
AI System: [System Name]
Assessment Date: [Date]
═══════════════════════════════════════════════════════════════════════════════

RISK SUMMARY:
─────────────────────────────────────────────────────────────────────────────
🔴 Critical: [X] risks
🟠 High:     [X] risks
🟡 Medium:   [X] risks
🟢 Low:      [X] risks
Total:       [X] risks identified

RISK HEAT MAP:
─────────────────────────────────────────────────────────────────────────────
              │ Negligible │ Minor │ Moderate │ Major │ Severe │
──────────────┼────────────┼───────┼──────────┼───────┼────────┤
Almost Certain│            │       │    2     │   1   │        │
Likely        │            │   1   │    3     │   2   │        │
Possible      │     1      │   2   │    4     │   1   │        │
Unlikely      │     2      │   3   │    2     │       │        │
Rare          │     1      │   1   │          │       │        │

TOP RISKS:
─────────────────────────────────────────────────────────────────────────────
1. [RISK-001] Training Data Bias
   Category: Data | Severity: Critical | Probability: Likely
   Impact: Discriminatory outcomes affecting protected groups
   Mitigation: Bias testing, diverse training data, fairness metrics
   Owner: [Name] | Due: [Date]

2. [RISK-002] Model Explainability
   Category: Technical | Severity: High | Probability: Possible
   Impact: Inability to explain decisions to regulators/users
   Mitigation: XAI techniques, model documentation, decision logs
   Owner: [Name] | Due: [Date]

[Continue for all high/critical risks...]

RISK REGISTER:
[Full risk register in table format]

MITIGATION PLAN:
[Prioritized mitigation actions]

═══════════════════════════════════════════════════════════════════════════════

---

Skill 4: Bias Testing Framework (/aig-bias-test)

Purpose

Design and execute bias testing for AI/ML models across protected characteristics.

Protected Characteristics

Characteristic	Examples	Regulations
Race/Ethnicity	Skin color, national origin	Civil Rights, EU AI Act
Gender	Sex, gender identity	Gender equality laws
Age	Age groups	Age discrimination laws
Disability	Physical, mental disabilities	ADA, accessibility laws
Religion	Religious beliefs	Religious freedom laws
Socioeconomic	Income, education level	Fair lending, housing

Bias Metrics

Metric	Formula	Target
Demographic Parity	P(Ŷ=1|A=0) = P(Ŷ=1|A=1)	Ratio 0.8-1.2
Equalized Odds	TPR and FPR equal across groups	Ratio 0.8-1.2
Predictive Parity	PPV equal across groups	Ratio 0.8-1.2
Calibration	P(Y=1|Ŷ=s) equal across groups	Difference <5%
Individual Fairness	Similar individuals → similar outcomes	Distance metric

Bias Testing Process

1. Define Protected Attributes
   └── Identify characteristics to test

2. Prepare Test Data
   └── Ensure representative samples for each group

3. Run Baseline Predictions
   └── Generate predictions for all groups

4. Calculate Fairness Metrics
   └── Compute metrics across groups

5. Identify Disparities
   └── Flag metrics outside acceptable range

6. Root Cause Analysis
   └── Investigate sources of bias

7. Mitigation Recommendations
   └── Propose corrections

8. Document Results
   └── Create bias testing report

Output Format

═══════════════════════════════════════════════════════════════════════════════
AI BIAS TESTING REPORT
Model: [Model Name]
Version: [Version]
Test Date: [Date]
═══════════════════════════════════════════════════════════════════════════════

OVERALL FAIRNESS SCORE: [XX]% - [PASS/FAIL/WARNING]

PROTECTED CHARACTERISTICS TESTED:
─────────────────────────────────────────────────────────────────────────────
☑ Gender (Male/Female)
☑ Age Groups (18-30, 31-50, 51-65, 65+)
☑ Ethnicity (Groups A, B, C, D)
☐ Disability (Not tested - data unavailable)

FAIRNESS METRICS:
─────────────────────────────────────────────────────────────────────────────
                    │ Gender │ Age    │ Ethnicity │ Target │ Status │
────────────────────┼────────┼────────┼───────────┼────────┼────────┤
Demographic Parity  │  0.92  │  0.85  │   0.78*   │ 0.8-1.2│ ⚠ WARN │
Equalized Odds (TPR)│  0.95  │  0.88  │   0.82    │ 0.8-1.2│ ✓ PASS │
Equalized Odds (FPR)│  0.91  │  0.79* │   0.84    │ 0.8-1.2│ ⚠ WARN │
Predictive Parity   │  0.98  │  0.92  │   0.89    │ 0.8-1.2│ ✓ PASS │

* Values outside acceptable range

DISPARITIES IDENTIFIED:
─────────────────────────────────────────────────────────────────────────────
1. [DISP-001] Ethnicity Group D has 22% lower positive prediction rate
   Metric: Demographic Parity = 0.78
   Root Cause: Under-representation in training data (3% vs 15% population)
   Recommendation: Augment training data, apply reweighting

2. [DISP-002] Age 65+ has higher false positive rate
   Metric: Equalized Odds (FPR) = 0.79
   Root Cause: Feature correlation with age-related variables
   Recommendation: Feature engineering, threshold adjustment

MITIGATION RECOMMENDATIONS:
─────────────────────────────────────────────────────────────────────────────
Priority 1: Address ethnicity disparity before production deployment
Priority 2: Monitor age-related FPR in production
Priority 3: Implement ongoing fairness monitoring

═══════════════════════════════════════════════════════════════════════════════

---

Skill 5: AI Policy Generator (/aig-policy)

Purpose

Generate comprehensive AI governance policies tailored to organization needs.

Policy Templates

Policy	Description	Audience
AI Governance Policy	Overall AI governance framework	All employees
AI Ethics Policy	Responsible AI principles	AI developers, users
AI Risk Management Policy	Risk identification and mitigation	Risk, compliance
AI Data Policy	Data use for AI systems	Data teams, AI teams
AI Vendor Policy	Third-party AI procurement	Procurement, legal
AI Acceptable Use Policy	Rules for using AI tools	End users
AI Model Documentation Policy	Model documentation requirements	AI developers
AI Incident Response Policy	Handling AI-related incidents	IT, security

Policy Structure

# [POLICY NAME]

## 1. Purpose
[Why this policy exists]

## 2. Scope
[Who and what this policy covers]

## 3. Definitions
[Key terms defined]

## 4. Policy Statements
### 4.1 [Principle 1]
[Policy details]

### 4.2 [Principle 2]
[Policy details]

## 5. Roles and Responsibilities
[Who is responsible for what]

## 6. Compliance
[How compliance is measured]

## 7. Exceptions
[Exception process]

## 8. Related Documents
[Links to related policies]

## 9. Review and Updates
[Review schedule]

## Document Control
- Version: [X.X]
- Effective Date: [Date]
- Owner: [Role]
- Approved By: [Name/Role]
- Next Review: [Date]

---

Skill 6: AI Ethics Framework (/aig-ethics)

Purpose

Develop responsible AI guidelines based on established ethical principles.

Core AI Ethics Principles

Principle	Description	Implementation
Fairness	AI treats all individuals equitably	Bias testing, diverse data
Transparency	AI decisions are explainable	XAI, documentation
Accountability	Clear ownership of AI outcomes	Governance structure
Privacy	Respect for personal data	Privacy by design
Safety	AI operates safely and reliably	Testing, monitoring
Human Oversight	Humans retain control	HITL, override capabilities
Beneficence	AI creates positive impact	Impact assessment

Ethics Review Checklist

□ Has the AI system been tested for bias?
□ Can the AI's decisions be explained to affected individuals?
□ Is there a clear owner accountable for the AI's outcomes?
□ Does the AI comply with privacy regulations?
□ Has the AI been tested for safety and reliability?
□ Can humans override or intervene in AI decisions?
□ Has the societal impact been assessed?
□ Are there feedback mechanisms for affected stakeholders?
□ Is there a process for ongoing monitoring?
□ Has the AI been reviewed by an ethics committee?

---

Skill 7: Model Risk Management (/aig-mrm)

Purpose

Implement model risk management framework for AI/ML models.

Model Risk Management Framework

┌─────────────────────────────────────────────────────────────────────────────┐
│                      MODEL RISK MANAGEMENT LIFECYCLE                        │
├─────────────────────────────────────────────────────────────────────────────┤
│                                                                             │
│  ┌──────────┐   ┌──────────┐   ┌──────────┐   ┌──────────┐   ┌──────────┐ │
│  │ Inventory│ → │ Tiering  │ → │ Validation│ → │ Approval │ → │ Monitoring│ │
│  └──────────┘   └──────────┘   └──────────┘   └──────────┘   └──────────┘ │
│       ↑                                                            │        │
│       └────────────────────────────────────────────────────────────┘        │
│                              Continuous Review                              │
│                                                                             │
└─────────────────────────────────────────────────────────────────────────────┘

Model Tiering

Tier	Criteria	Validation Rigor	Review Frequency
Tier 1 (High)	Material financial impact, regulatory	Full independent validation	Annual
Tier 2 (Medium)	Moderate impact, operational	Focused validation	18 months
Tier 3 (Low)	Low impact, internal use	Self-assessment	24 months

Model Documentation (Model Card)

# Model Card: [Model Name]

## Model Details
- Name: [Name]
- Version: [X.X]
- Type: [Classification/Regression/etc.]
- Owner: [Team/Individual]
- Tier: [1/2/3]

## Intended Use
- Primary use case: [Description]
- Users: [Who uses this model]
- Out-of-scope uses: [What it shouldn't be used for]

## Training Data
- Source: [Data sources]
- Size: [Number of records]
- Date range: [Time period]
- Known limitations: [Data limitations]

## Performance Metrics
- Accuracy: [XX%]
- Precision: [XX%]
- Recall: [XX%]
- AUC-ROC: [X.XX]

## Fairness Metrics
- [Metric 1]: [Value]
- [Metric 2]: [Value]

## Limitations & Risks
- [Limitation 1]
- [Limitation 2]
- [Risk 1]

## Monitoring
- Drift detection: [Method]
- Performance threshold: [Criteria for retraining]
- Alert contacts: [Who to notify]

---

Skill 8: AI Inventory Manager (/aig-inventory)

Purpose

Catalog and classify all AI systems within the organization.

AI System Classification

By Risk Level (EU AI Act)

Risk Level	Examples	Requirements
Unacceptable	Social scoring, real-time biometric ID	Prohibited
High	Credit scoring, recruitment, medical devices	Full compliance
Limited	Chatbots, emotion recognition	Transparency
Minimal	Spam filters, recommendation systems	No requirements

By Use Case

Category	Examples
Customer-Facing	Chatbots, recommendation engines, pricing
Internal Operations	Forecasting, automation, analytics
Decision Support	Risk assessment, fraud detection
Autonomous	Self-driving, trading algorithms

AI Inventory Template

{
  "ai_system_id": "AI-001",
  "name": "Customer Churn Prediction Model",
  "description": "Predicts customer churn probability",
  "status": "production",
  "risk_level": "high",
  "eu_ai_act_classification": "high_risk",
  "business_owner": "Marketing Director",
  "technical_owner": "Data Science Lead",
  "deployment_date": "2025-06-15",
  "last_validation": "2026-01-10",
  "next_review": "2026-07-10",
  "model_type": "gradient_boosting",
  "training_data": "Customer transactions 2020-2025",
  "inputs": ["transaction_history", "demographics", "engagement"],
  "outputs": ["churn_probability", "risk_segment"],
  "users": ["marketing_team", "retention_team"],
  "integrations": ["crm_system", "marketing_automation"],
  "documentation": "/docs/models/churn_model_v2.md",
  "bias_testing": {
    "last_test": "2026-01-10",
    "result": "pass",
    "next_test": "2026-04-10"
  },
  "monitoring": {
    "drift_detection": true,
    "performance_alerts": true,
    "threshold": "accuracy < 0.85"
  }
}

Inventory Dashboard Output

═══════════════════════════════════════════════════════════════════════════════
AI SYSTEMS INVENTORY
Organization: [Name]
As of: [Date]
═══════════════════════════════════════════════════════════════════════════════

SUMMARY:
─────────────────────────────────────────────────────────────────────────────
Total AI Systems: [XX]
├── Production:    [XX]
├── Development:   [XX]
├── Retired:       [XX]
└── Planned:       [XX]

BY RISK LEVEL:
─────────────────────────────────────────────────────────────────────────────
🔴 High Risk:      [XX] systems
🟡 Limited Risk:   [XX] systems
🟢 Minimal Risk:   [XX] systems

BY STATUS:
─────────────────────────────────────────────────────────────────────────────
✅ Compliant:      [XX] systems
⚠️ Review Due:     [XX] systems
❌ Non-Compliant:  [XX] systems

ATTENTION REQUIRED:
─────────────────────────────────────────────────────────────────────────────
1. [AI-003] Credit Scoring Model - Validation overdue by 30 days
2. [AI-007] Fraud Detection - Bias test failed, remediation needed
3. [AI-012] Recruitment Screener - Documentation incomplete

INVENTORY LIST:
[Full table of all AI systems]

═══════════════════════════════════════════════════════════════════════════════

---

Key Frameworks Referenced

Framework	Description	Link
ISO/IEC 42001:2023	AI Management System	ISO
NIST AI RMF	AI Risk Management Framework	NIST
EU AI Act	European AI Regulation	EU
IEEE 7000	Ethical AI Design	IEEE
OECD AI Principles	International AI Guidelines	OECD

---

Deliverable Templates

Deliverable	Format	Template Location
AI Readiness Assessment	DOCX, PDF	templates/aig-readiness-report.md
ISO 42001 Gap Analysis	XLSX, PDF	templates/aig-iso42001-gap.md
AI Risk Register	XLSX	templates/aig-risk-register.xlsx
Bias Testing Report	PDF	templates/aig-bias-report.md
AI Governance Policy	DOCX	templates/aig-policy.md
AI Ethics Framework	DOCX	templates/aig-ethics.md
Model Documentation	MD	templates/aig-model-card.md
AI Inventory	XLSX	templates/aig-inventory.xlsx

---

Integration with Other Skills

Skill	Integration
/cps-budget	Fee calculation for AIG engagements
/doc-gen	Generate formatted deliverables
/proposal	AIG-specific proposal content
/iso-42001	Detailed ISO 42001 implementation

---

Service Line: AIG (AI Governance & Management) Version: 1.0 Last Updated: 2026-02-02

---

Additional Skill: ai-incident-response

Command: /cps-skills:ai-incident-response (v4.2.1)

Purpose

Design AI incident response playbook for model failures, bias incidents, hallucinations, abuse.

Inputs required

client:
  name: "Client Name"
context:
  scope: "in-scope description"
  constraints: ["regulatory / commercial constraints"]

Methodology

1. Frame the request against AI Governance domain conventions.
2. Pull required inputs from engagement brain and external sources.
3. Apply the ai-incident-response method using the relevant subagent.
4. Synthesize into deliverable with recommendations + risks.
5. Validate via /cps:verify-quality.

Output shape

CPS-branded ai-incident-response deliverable in 05_Deliverables_Final/.

Quality checklist

• Pyramid Principle structure
• Source citations on all data
• Quantified impact where the analysis supports it

---

Additional Skill: ai-bias-audit

Command: /cps-skills:ai-bias-audit (v4.2.1)

Purpose

Audit AI system for protected-class bias using fairness metrics (statistical parity, equal opportunity, etc.).

Inputs required

client:
  name: "Client Name"
context:
  scope: "in-scope description"
  constraints: ["regulatory / commercial constraints"]

Methodology

1. Frame the request against AI Governance domain conventions.
2. Pull required inputs from engagement brain and external sources.
3. Apply the ai-bias-audit method using the relevant subagent.
4. Synthesize into deliverable with recommendations + risks.
5. Validate via /cps:verify-quality.

Output shape

CPS-branded ai-bias-audit deliverable in 05_Deliverables_Final/.

Quality checklist

• Pyramid Principle structure
• Source citations on all data
• Quantified impact where the analysis supports it