Secure SDLC
Security activities are mapped to each development phase, informed by the NIST Secure Software Development Framework.
- Security requirements defined with functional scope
- Design review before implementation
- Verification before release
- Post-release monitoring and remediation
Code Reviews
Changes are reviewed before they reach a protected branch, with attention to authentication, authorization, input handling, and data exposure.
- Peer review required for protected branches
- Security-focused review checklist
- Rejection of unreviewed direct pushes
- Traceability from change to reviewer
Threat Modeling
Systems handling sensitive data or privileged operations are modeled for attacker goals, entry points, and abuse cases before build.
- Trust boundary identification
- Abuse case enumeration
- MITRE ATT&CK informed technique mapping
- Mitigations tracked as work items
Dependency Scanning
Third-party packages are inventoried and evaluated for known vulnerabilities, with remediation prioritized by exploitability and exposure.
- Automated dependency vulnerability checks
- Prioritization by severity and reachability
- Upgrade or replacement decisions recorded
- Transitive dependency visibility
Secret Detection
Credentials are kept out of source control. Automated detection runs against changes, and any exposure triggers rotation.
- Automated secret scanning on changes
- Managed secret storage for runtime values
- Rotation on suspected exposure
- No credentials in client bundles
Static Analysis
Static analysis and type checking run in the pipeline to catch insecure patterns before deployment.
- Static analysis in continuous integration
- Strict type checking
- Lint rules for unsafe patterns
- Findings triaged rather than suppressed by default
Dynamic Analysis
Running applications are exercised against OWASP Top 10 and API Security Top 10 categories in a controlled environment.
- Authenticated and unauthenticated test passes
- Input validation and access control testing
- Authorization boundary testing
- Findings tracked to closure
Secure CI/CD
Pipelines are treated as production infrastructure: least-privilege credentials, protected branches, and auditable deployments.
- Scoped deployment credentials
- Protected branches and required checks
- Immutable, traceable build artifacts
- Audit trail for every deployment
Infrastructure as Code
Infrastructure is defined declaratively so configuration is reviewable, repeatable, and comparable against a baseline.
- Reviewed infrastructure definitions
- Repeatable environment provisioning
- Configuration drift detection
- Baseline standards applied consistently
Supply Chain Security
Build inputs are verified, pinned where practical, and sourced from trusted registries.
- Trusted registry sources
- Version pinning and lockfiles
- Provenance review for critical components
- Patch verification before promotion
AI-Assisted Secure Coding
AI assistance is used under human review. Generated code is subject to the same review, scanning, and testing as any other change, and sensitive data is not submitted to third-party models without authorization.
- Human review required for AI-generated code
- Same scanning and testing gates apply
- No sensitive or customer data in prompts without authorization
- AI usage governed by the AI governance policy