Threat Modeling Framework Skill

Context

This skill applies when:

• Designing new features or system components
• Conducting security architecture reviews
• Documenting threat models for SDLC phases
• Identifying security threats using STRIDE methodology
• Creating threat model evidence portfolios
• Integrating threat modeling into CI/CD pipelines
• Reviewing threat models during PR process

This skill enforces Hack23 Secure Development Policy Section 🕷️ requirements for systematic threat analysis.

Rules

1. STRIDE Threat Modeling (Policy Section 🕷️.1)

1. Spoofing: Identify authentication and identity verification threats
2. Tampering: Identify data integrity and unauthorized modification threats
3. Repudiation: Identify logging, audit trail, and non-repudiation threats
4. Information Disclosure: Identify confidentiality and data exposure threats
5. Denial of Service: Identify availability and resource exhaustion threats
6. Elevation of Privilege: Identify authorization and privilege escalation threats

2. Threat Model Documentation (Policy Section 🕷️.1.2)

7. System Diagram: Data flow diagrams with trust boundaries
8. Asset Identification: Critical assets and their classification
9. Threat Enumeration: STRIDE threats for each component
10. Mitigation Strategies: Security controls for each identified threat
11. Risk Assessment: Likelihood and impact ratings
12. Evidence Links: References to implemented controls in code

3. SDLC Integration (Policy Section 🕷️.1.3)

13. Planning Phase: Threat model before design finalization
14. Development Phase: Update threat model as design evolves
15. Testing Phase: Verify mitigations are implemented
16. Release Phase: Threat model sign-off required
17. Maintenance Phase: Update threat model for changes

4. Evidence Portfolio (Policy Section 🕷️.2)

18. THREAT_MODEL.md: Structured threat model documentation
19. Architecture Diagrams: Mermaid diagrams showing trust boundaries
20. Mitigation Evidence: Links to code implementing controls
21. Test Evidence: Links to security tests validating mitigations
22. Review Records: PR approvals and security review sign-offs

Examples

✅ Good Pattern: STRIDE Threat Model for MCP Tool

# Threat Model: search_meps MCP Tool

**Component**: European Parliament MCP Server - search_meps tool  
**Classification**: Public data with personal information (MEP contact details)  
**Last Updated**: 2026-02-16  
**Reviewed By**: Security Team

## System Overview

```mermaid
graph LR
    Client[MCP Client<br/>Untrusted] -->|JSON-RPC| Server[MCP Server<br/>Trusted]
    Server -->|HTTPS| EPAPI[EP API<br/>External]
    Server --> Cache[LRU Cache<br/>Trusted]
    
    style Client fill:#ff6b6b
    style Server fill:#51cf66
    style EPAPI fill:#ffd43b
    style Cache fill:#51cf66

Assets

1. MEP Personal Data - Names, countries, email addresses (Confidentiality: Medium, Integrity: High)
2. API Credentials - European Parliament API access (Confidentiality: High)
3. Cache Data - Temporary storage of MEP data (Confidentiality: Medium)
4. MCP Server Availability - Service uptime (Availability: High)

STRIDE Analysis

Spoofing Identity (S)

Risk Rating: LOW (mitigated)

Tampering (T)

Risk Rating: LOW (mitigated)

Repudiation (R)

Risk Rating: LOW (mitigated)

Information Disclosure (I)

Risk Rating: MEDIUM (monitor cache TTL compliance)

Denial of Service (D)

Risk Rating: MEDIUM (monitor resource usage)

Elevation of Privilege (E)

Risk Rating: LOW (mitigated)

Residual Risks

1. GDPR Compliance: Cache TTL must be monitored to ensure 24-hour limit
2. Resource Exhaustion: Monitor memory and CPU usage under load
3. EP API Changes: Breaking changes to EP API could bypass validation

Review & Approval

• Threat model reviewed by Security Team
• Mitigations implemented and verified
• Security tests passing (80% coverage)
• GDPR compliance validated
• Ready for production deployment

Approved By: Security Team
Date: 2026-02-16

**Policy Reference**: [Secure Development Policy Section 🕷️.1](https://github.com/Hack23/ISMS-PUBLIC/blob/main/Secure_Development_Policy.md#threat-modeling-requirements)

**Evidence**: [CIA Threat Model Example](https://github.com/Hack23/cia/blob/master/THREAT_MODEL.md)

### ✅ Good Pattern: Trust Boundary Diagram

```mermaid
graph TB
    subgraph "Untrusted Zone"
        Client[MCP Client]
        User[End User]
    end
    
    subgraph "Trusted Zone - MCP Server"
        InputVal[Input Validation<br/>Zod Schemas]
        Tools[Tool Handlers]
        Cache[LRU Cache]
        Audit[Audit Logger]
    end
    
    subgraph "External Zone"
        EPAPI[European Parliament API<br/>HTTPS Only]
    end
    
    User -->|User Input| Client
    Client -->|JSON-RPC<br/>stdio| InputVal
    InputVal -->|Validated| Tools
    Tools -->|Store| Cache
    Tools -->|Log Access| Audit
    Tools -->|HTTPS Request| EPAPI
    EPAPI -->|JSON Response| Tools
    Tools -->|MCP Response| Client
    
    style Client fill:#ff6b6b
    style User fill:#ff6b6b
    style InputVal fill:#51cf66
    style Tools fill:#51cf66
    style Cache fill:#51cf66
    style Audit fill:#51cf66
    style EPAPI fill:#ffd43b

Trust Boundaries:

• Red (Untrusted): MCP Client, End User - all input is potentially malicious
• Green (Trusted): MCP Server components - validated and authorized
• Yellow (External): European Parliament API - third-party, HTTPS required

✅ Good Pattern: Threat Model Integration in PR Template

## Security Checklist

### Threat Model Updates
- [ ] Threat model reviewed for this change
- [ ] New threats identified and documented
- [ ] Mitigations implemented for new threats
- [ ] STRIDE analysis updated if architecture changed
- [ ] Trust boundaries reviewed and validated

### Evidence Links
- Threat Model: [THREAT_MODEL.md](link)
- Architecture Diagram: [ARCHITECTURE.md#security-architecture](link)
- Security Tests: [tests/security/](link)
- GDPR Compliance: [.github/skills/gdpr-compliance/](link)

### Sign-off
- [ ] Security team review completed
- [ ] Threat model approved
- [ ] All mitigations verified in code
- [ ] Security tests passing

Policy Reference: Secure Development Policy Section 🕷️.1.3

✅ Good Pattern: Automated Threat Model Validation

/**
 * Threat model validation script
 * Ensures threat model documentation is complete and up-to-date
 * 
 * ISMS Policy: Secure Development Policy Section 🕷️.1.2
 * Evidence: https://github.com/Hack23/ISMS-PUBLIC/blob/main/Secure_Development_Policy.md#required-threat-model-documentation
 */

interface ThreatModel {
  component: string;
  lastUpdated: string;
  reviewedBy: string;
  systemDiagram: boolean;
  assets: Array<{
    name: string;
    classification: string;
    cia: { confidentiality: string; integrity: string; availability: string };
  }>;
  threats: {
    spoofing: Array<{ id: string; threat: string; mitigation: string; evidence: string }>;
    tampering: Array<{ id: string; threat: string; mitigation: string; evidence: string }>;
    repudiation: Array<{ id: string; threat: string; mitigation: string; evidence: string }>;
    informationDisclosure: Array<{ id: string; threat: string; mitigation: string; evidence: string }>;
    denialOfService: Array<{ id: string; threat: string; mitigation: string; evidence: string }>;
    elevationOfPrivilege: Array<{ id: string; threat: string; mitigation: string; evidence: string }>;
  };
  residualRisks: string[];
  approved: boolean;
  approvedBy?: string;
  approvalDate?: string;
}

function validateThreatModel(threatModel: ThreatModel): { valid: boolean; errors: string[] } {
  const errors: string[] = [];
  
  // Check required fields
  if (!threatModel.component) {
    errors.push('Missing component name');
  }
  
  if (!threatModel.systemDiagram) {
    errors.push('System diagram required per Policy Section 🕷️.1.2');
  }
  
  // Check asset identification
  if (threatModel.assets.length === 0) {
    errors.push('At least one asset must be identified');
  }
  
  // Check STRIDE coverage
  const strideCategories = Object.keys(threatModel.threats);
  if (strideCategories.length !== 6) {
    errors.push('All 6 STRIDE categories must be analyzed');
  }
  
  // Check each threat has mitigation and evidence
  for (const [category, threats] of Object.entries(threatModel.threats)) {
    for (const threat of threats) {
      if (!threat.mitigation) {
        errors.push(`${category}:${threat.id} missing mitigation`);
      }
      if (!threat.evidence) {
        errors.push(`${category}:${threat.id} missing evidence link`);
      }
    }
  }
  
  // Check approval
  if (!threatModel.approved) {
    errors.push('Threat model must be approved before release');
  }
  
  // Check review date (must be within 90 days)
  const lastUpdated = new Date(threatModel.lastUpdated);
  const now = new Date();
  const daysSinceUpdate = (now.getTime() - lastUpdated.getTime()) / (1000 * 60 * 60 * 24);
  
  if (daysSinceUpdate > 90) {
    errors.push(`Threat model outdated (${Math.floor(daysSinceUpdate)} days). Update required per Policy Section 🕷️.1.3`);
  }
  
  return {
    valid: errors.length === 0,
    errors,
  };
}

// CI/CD integration
async function checkThreatModelInCI(): Promise<void> {
  const threatModelPath = './THREAT_MODEL.json';
  
  if (!fs.existsSync(threatModelPath)) {
    console.error('❌ THREAT_MODEL.json not found');
    console.error('Required per Secure Development Policy Section 🕷️.1.2');
    console.error('See: https://github.com/Hack23/ISMS-PUBLIC/blob/main/Secure_Development_Policy.md#required-threat-model-documentation');
    process.exit(1);
  }
  
  const threatModel: ThreatModel = JSON.parse(fs.readFileSync(threatModelPath, 'utf-8'));
  const validation = validateThreatModel(threatModel);
  
  if (!validation.valid) {
    console.error('❌ Threat model validation failed:');
    validation.errors.forEach(err => console.error(`  - ${err}`));
    process.exit(1);
  }
  
  console.log('✅ Threat model validated successfully');
}

Evidence: Black Trigram Threat Model CI

Anti-Patterns

❌ Bad: No Threat Model

// Just wrote code without thinking about threats!
export async function handleSensitiveData(data: any) {
  return await processData(data);
}

Why: Violates Secure Development Policy Section 🕷️ - All features require threat modeling

❌ Bad: Incomplete STRIDE Analysis

# Threat Model

## Threats
- Someone might hack the system

## Mitigations
- Use encryption

Why: Not following STRIDE methodology, no specific threats, no evidence links

Evidence Portfolio

Reference Implementations

1. Citizen Intelligence Agency (CIA)
   • Threat Model: https://github.com/Hack23/cia/blob/master/THREAT_MODEL.md
   • Architecture Security: https://github.com/Hack23/cia/blob/master/SECURITY_ARCHITECTURE.md
2. Black Trigram Game
   • Threat Model: https://github.com/Hack23/blacktrigram/blob/main/THREAT_MODEL.md
   • CI Validation: https://github.com/Hack23/blacktrigram/blob/main/.github/workflows/threat-model-check.yml
3. CIA Compliance Manager
   • Compliance Threat Model: https://github.com/Hack23/cia-compliance-manager/blob/main/THREAT_MODEL.md

Policy Documents

• Secure Development Policy Section 🕷️: https://github.com/Hack23/ISMS-PUBLIC/blob/main/Secure_Development_Policy.md#advanced-security-testing-framework
• STRIDE Methodology: https://learn.microsoft.com/en-us/azure/security/develop/threat-modeling-tool-threats

ISMS Compliance

This skill enforces:

• SD-TM-001: STRIDE threat modelling methodology
• SD-TM-002: Threat model documentation requirements
• SD-TM-003: SDLC integration — model at design / before implementation
• SD-TM-004: Evidence portfolio maintenance — linked to ADRs and architecture docs

Policy References

Primary:

• Threat Modeling
• Secure Development Policy (Section 🕷️ Advanced Security Testing Framework)

Related:

• Information Security Policy
• Vulnerability Management — Threats → risks → remediation issues
• Change Management — Revisit model on significant change
• Privacy Policy — GDPR-specific threats
• OWASP LLM Security Policy — Prompt-injection STRIDE branch
• Open Source Policy — Supply-chain threats