Secure Development Lifecycle

🛡️ Secure Development Lifecycle (SDLC) Skill

🔴 AI FIRST Quality Principle

Apply the AI FIRST principle: never accept first-pass quality. Minimum 2 iterations. Read all output, improve every section. No shortcuts.

🎯 Purpose

Comprehensive security practices for the entire Software Development Lifecycle (SDLC), ensuring security is built in from inception through maintenance. Integrates classification-driven requirements, AI-augmented development controls, and systematic testing frameworks aligned with Hack23 Secure Development Policy.

🔗 Governance chain

Authority flows from the Information Security Policy (master ISMS document) → Secure Development Policy (SDLC requirements) → this skill (operational implementation in this repository).

Directly relevant companion policies:

• Open Source Policy — every dependency introduced during SDLC must be licence-cleared and tracked in SBOM
• AI Policy — Copilot and agent-generated code require human review and DCO sign-off
• Threat Modeling · Vulnerability Management · Change Management · Access Control · Cryptography · Incident Response · Security Metrics · CLASSIFICATION

🔐 Core Security Principles

🔐 Security by Design

• 🏷️ Project Classification: CIA triad, RTO/RPO, business impact analysis
• 🛡️ Secure Coding Standards: OWASP Top 10 alignment with classification controls
• 🏗️ Architecture Documentation: SECURITY_ARCHITECTURE.md + FUTURE_SECURITY_ARCHITECTURE.md

🌟 Transparency Through Documentation

• 📋 Living Security Architecture: Real-time documentation with classification controls
• 🎖️ Public Security Badges: OpenSSF Scorecard, SLSA, Quality Gate validation
• 🔓 Open Development: Demonstrating expertise while maintaining classification

🔄 Continuous Security Improvement

• 🏷️ Classification-Driven Testing: SAST/SCA/DAST per classification levels
• 📈 Performance Monitoring: Security metrics with availability SLAs
• 🔍 Regular Reviews: Classification-based risk management and ROI

🔄 5-Phase SDLC Security Framework

📋 Phase 1: Planning & Design

🏷️ Project Classification (REQUIRED)

Apply Classification Framework:

• CIA Triad Analysis (Confidentiality, Integrity, Availability)
• Business Impact Classification (Revenue, Trust, Compliance)
• RTO/RPO Definition (Recovery Time/Point Objectives)
• Risk Assessment Integration with Risk Register
• Cost-Benefit Analysis (Security ROI)

Classification Levels:

🏗️ Security Architecture Design (REQUIRED)

Maintain comprehensive architecture documentation:

• SECURITY_ARCHITECTURE.md: Current implemented security design
• FUTURE_SECURITY_ARCHITECTURE.md: Planned security improvements
• ARCHITECTURE.md: Complete C4 models (Context, Container, Component, Code)
• DATA_MODEL.md: Data structures and classifications
• FLOWCHART.md: Business process flows with security controls

🎯 Threat Modeling (MANDATORY)

Per Threat Modeling Policy:

• STRIDE Framework: Spoofing, Tampering, Repudiation, Information Disclosure, DoS, Elevation of Privilege
• MITRE ATT&CK Integration: 14 tactics mapped with techniques
• Attack Tree Analysis: Graphical attack path decomposition
• Threat Agent Classification: 7 categories (Accidental Insiders → Nation-State APTs)
• THREAT_MODEL.md: Comprehensive 9-section threat documentation

💻 Phase 2: Development

🛡️ Secure Coding Guidelines

OWASP Top 10 (2021) Alignment:

1. A01 - Broken Access Control: Proper authentication/authorization
2. A02 - Cryptographic Failures: TLS 1.3, AES-256 encryption
3. A03 - Injection: Parameterized queries, input validation
4. A04 - Insecure Design: Apply threat modeling, secure patterns
5. A05 - Security Misconfiguration: Secure defaults, hardened configs
6. A06 - Vulnerable Components: SCA scanning, SBOM generation
7. A07 - Authentication Failures: MFA, secure session management
8. A08 - Software/Data Integrity: Code signing, integrity checks
9. A09 - Logging Failures: Comprehensive security event logging
10. A10 - SSRF: Validate external resource requests

🔍 Code Review Requirements

Classification-Based Review:

🔐 Secret Management (MANDATORY)

• Zero Hard-Coded Credentials: No secrets in source code
• GitHub Secrets: All credentials in encrypted secrets
• Rotation Policy: Critical: 90 days, High: 180 days, Medium/Low: 365 days
• Access Logging: All secret access logged and monitored
• Least Privilege: Secrets scoped to minimum required access

🧪 Phase 3: Security Testing

🔬 Static Application Security Testing (SAST)

Implementation:

• Tool: SonarCloud integration on every commit
• Quality Gates: Classification-based failure thresholds
• Coverage: All code analyzed for security vulnerabilities
• Reporting: Public quality/security dashboards

Classification-Based Quality Gates:

📦 Software Composition Analysis (SCA)

Dependency Security:

• Automated Scanning: Dependabot, Snyk, or equivalent
• SBOM Generation: Software Bill of Materials for all releases
• Vulnerability Database: CVE, NVD, GitHub Advisory integration
• Update Policy: Classification-based patching SLAs
• License Compliance: OSS license validation

Remediation SLAs:

⚡ Dynamic Application Security Testing (DAST)

Runtime Security Testing:

• Tool: OWASP ZAP, Burp Suite, or equivalent
• Scope: Staging environments (classification-appropriate)
• Frequency: Per sprint (Critical/High), quarterly (Medium/Low)
• Coverage: All authentication, authorization, input handling paths

🔍 Secret Scanning (CONTINUOUS)

• GitHub Secret Scanning: Enabled on all repositories
• Pre-commit Hooks: Detect secrets before commit
• Historical Scanning: Scan entire git history
• Alert Integration: Immediate notifications to security team
• Remediation SLA: Critical secrets rotated within 1 hour

📋 Test Data Protection (MANDATORY)

• Zero Production Data: Never use real data in dev/test
• Data Anonymization: Pseudonymize test data
• Secure Deletion: Wipe test data after use
• Access Control: Least privilege for test environments

🎯 Unit Test Coverage & Quality

📊 Testing Standards

Minimum Thresholds:

• Line Coverage: ≥80% (Critical/High), ≥70% (Medium/Low)
• Branch Coverage: ≥70% (Critical/High), ≥60% (Medium/Low)
• Mutation Testing: ≥60% mutation score (Critical only)
• Test Execution: Every commit and PR
• Trend Analysis: Historical tracking, regression prevention

📚 Required Documentation

Every repository MUST have:

• UnitTestPlan.md: Comprehensive unit test strategy
• Test Results: Public HTML reports (GitHub Pages)
• Coverage Dashboards: Accessible coverage metrics
• Quality Badges: Status badges in README.md

📊 Reference Implementation Examples

🏛️ Citizen Intelligence Agency (Java/Spring):

🎮 Black Trigram (TypeScript/Phaser):

📊 CIA Compliance Manager (TypeScript/Vite):

🌐 End-to-End Testing Strategy

🎯 E2E Testing Requirements

Coverage Areas:

• Critical User Journeys: All primary workflows tested
• Authentication Flows: Login, logout, session management
• Authorization Checks: Role-based access validation
• Data Integrity: CRUD operations validation
• Performance: Response time within SLA thresholds

📚 Required Documentation

Every repository MUST have:

• E2ETestPlan.md: Comprehensive E2E test strategy
• Mochawesome Reports: Public HTML test results
• Browser Matrix: Cross-browser validation (Chrome, Firefox, Safari, Edge)
• Performance Assertions: Response time validation

📊 Reference Implementation Examples

🏛️ Citizen Intelligence Agency:

🎮 Black Trigram:

📊 CIA Compliance Manager:

🤖 AI-Augmented Development Controls

🔐 AI as Proposal Generator, Not Authority

Core Principles:

• All AI outputs are proposals: Require human review and approval
• No autonomous deployment: AI cannot bypass CI/CD pipelines or security gates
• Human accountability: Responsibility remains with human developers
• Transparent attribution: Document AI assistance in PR descriptions

📋 PR Review Requirements

Mandatory Controls:

• Human Review: All AI-assisted changes pass through standard PR workflows
• Security Gate Enforcement: CI pipelines unchanged or only tightened
• Change Attribution: PR descriptions MUST document AI tools used
• Code Ownership: Human developers remain code owners

🔧 Curator-Agent Configuration Management

Change Control:

• Scope: .github/agents/*.md, .github/copilot-mcp*.json, .github/workflows/copilot-setup-steps.yml
• Classification: Normal Change per Change Management
• Approval: CEO or designated security owner required
• Risk Assessment: Documented evaluation for capability expansion

🛡️ Security Requirements

Tool Governance:

• Least Privilege: Agents operate with minimal required tool access
• MCP Configuration Control: Model Context Protocol changes require security review
• Audit Trail: All agent activities logged for compliance analysis
• Capability Expansion: New integrations require documented risk assessment

🚀 Phase 4: Deployment

🤖 Automated CI/CD Pipelines

Security Gates:

• SAST Scanning: Code quality gates (classification-based thresholds)
• SCA Scanning: Dependency vulnerability checks with auto-block
• Secret Scanning: Zero tolerance for exposed credentials
• Container Scanning: Image vulnerability assessment (if applicable)
• Infrastructure as Code: Terraform/CloudFormation security validation

✅ Manual Approval Gates

Classification-Based Approvals:

📋 Deployment Checklists

Pre-Deployment Verification:

• All security tests passing
• Classification-appropriate controls validated
• Rollback plan documented
• Monitoring alerts configured
• Incident response procedures ready

📊 Security Metrics

Real-Time Monitoring:

• OpenSSF Scorecard: Public security posture metrics
• SLSA Level: Supply chain security attestation
• Quality Gates: SonarCloud quality/security dashboards
• Uptime Metrics: Availability aligned with classification SLAs

🔧 Phase 5: Maintenance & Operations

🆘 Vulnerability Management

Classification-Based Remediation: Per Vulnerability Management:

📈 Performance Monitoring

Security Metrics Integration: Per Security Metrics:

• Availability Tracking: Uptime per classification requirements
• Response Time: Performance within SLA thresholds
• Error Rates: Security-relevant errors logged and analyzed
• Incident Metrics: MTTR, MTTD aligned with classification

🔄 Regular Updates

Patch Management:

• Security Patches: Classification-based deployment schedules
• Dependency Updates: Automated PRs with security review
• Framework Updates: Major version upgrades with testing
• Business Continuity: Updates aligned with BCP

📋 Incident Response

Integration: Per Incident Response Plan:

• Classification-Driven Escalation: Incident severity based on project classification
• Communication Procedures: Stakeholder notifications per classification
• Recovery Objectives: RTO/RPO aligned with classification
• Post-Incident Review: Lessons learned and improvement actions

📊 SDLC Security Maturity Levels

Level 1: Basic (Minimum Viable Security)

• ✅ Basic security controls implemented
• ✅ Dependabot enabled
• ✅ Secret scanning active
• ✅ Basic threat model documented

Level 2: Intermediate (Standard Security)

• ✅ Level 1 + Classification implemented
• ✅ SAST/SCA integrated in CI/CD
• ✅ Unit test coverage ≥70%
• ✅ SECURITY_ARCHITECTURE.md maintained
• ✅ Regular vulnerability scanning

Level 3: Advanced (Enhanced Security)

• ✅ Level 2 + DAST implementation
• ✅ Comprehensive threat modeling (STRIDE + MITRE ATT&CK)
• ✅ Unit test coverage ≥80%
• ✅ E2E testing framework
• ✅ Public security dashboards

Level 4: Mature (Security Excellence)

• ✅ Level 3 + AI-augmented development controls
• ✅ Mutation testing (≥60% score)
• ✅ Full C4 architecture documentation
• ✅ Continuous security monitoring
• ✅ Evidence-based compliance (badges, reports)
• ✅ External security validation (pentesting, audits)

✅ SDLC Security Checklist

Planning & Design Phase

• Project classification completed (CIA triad, RTO/RPO, business impact)
• Threat model documented (STRIDE + MITRE ATT&CK)
• Security architecture designed (C4 models, data flows)
• Risk assessment integrated with Risk Register
• Cost-benefit analysis for security investments

Development Phase

• Secure coding standards applied (OWASP Top 10)
• Code review requirements met (classification-based)
• Asset classification implemented
• Secret management controls enforced
• AI-augmented development controls active

Testing Phase

• SAST scanning integrated (SonarCloud)
• SCA scanning enabled (Dependabot)
• DAST testing implemented (OWASP ZAP)
• Secret scanning active (GitHub)
• Unit test coverage thresholds met (≥80% line, ≥70% branch)
• E2E testing framework operational
• Test data protection controls enforced

Deployment Phase

• CI/CD security gates configured
• Manual approval gates per classification
• Deployment checklists completed
• Security metrics monitoring active
• Rollback procedures documented

Maintenance Phase

• Vulnerability management process active
• Performance monitoring with security metrics
• Regular update schedule defined
• Incident response procedures integrated
• Continuous improvement process operational

📚 References

Hack23 ISMS Core Policies

• 🛠️ Secure Development Policy - Comprehensive SDLC framework
• 🏷️ Classification Framework - Business impact analysis
• 🎯 Threat Modeling Policy - Systematic threat analysis
• 📉 Risk Register - Enterprise risk management
• 🔍 Vulnerability Management - Remediation procedures
• 📊 Security Metrics - KPI tracking
• 🚨 Incident Response Plan - Security incident procedures
• 🔄 Business Continuity Plan - BCP/DR processes
• 📝 Change Management - Change control procedures
• 🏷️ Data Classification Policy - Data handling requirements

Example Implementations

• 🏛️ CIA Security Architecture - Full authentication stack (Java/Spring)
• 🏛️ CIA Threat Model - Comprehensive threat analysis
• 📊 CIA Compliance Manager Security - Frontend security (TypeScript/Vite)
• 🎮 Black Trigram Security - Gaming security (TypeScript/Phaser)
• 🗳️ Riksdagsmonitor Security - Static site security (HTML/CSS)

External Frameworks

• OWASP Top 10 - Critical web application security risks
• OWASP ASVS - Application security verification
• NIST SP 800-218 - Secure Software Development Framework
• Microsoft SDL - Security Development Lifecycle
• MITRE ATT&CK - Adversary tactics and techniques

🎯 Remember

• Classification Drives Security: All requirements aligned with business impact
• Transparency is Competitive Advantage: Public security demonstrates expertise
• AI Augments, Humans Decide: AI proposals require human approval
• Evidence-Based Security: Badges, dashboards, reports validate claims
• Continuous Improvement: Measure, analyze, improve security posture
• Documentation is Mandatory: SECURITY_ARCHITECTURE.md, THREAT_MODEL.md required
• Testing is Not Optional: Unit + E2E coverage proves quality
• Security is Everyone's Responsibility: DevSecOps culture required

Last Updated: 2026-02-10 (Continuous)
Version: Based on Hack23 Secure Development Policy v2.1 & STYLE_GUIDE v2.3