The SDD plugin is built on the IEEE SWEBOK v4 (Software Engineering Body of Knowledge) and follows established industry standards for requirements engineering, design, construction, testing, and quality management.
SWEBOK v4 alignment
The Software Engineering Body of Knowledge (SWEBOK) defines the foundational knowledge areas of software engineering. The SDD plugin implements SWEBOK principles across all pipeline stages.
Chapter 1: Software Requirements
Pipeline stage: Requirements Engineering (/sdd:requirements-engineer)
SWEBOK topics covered:
| Topic | Implementation |
|---|
| Requirements Fundamentals | EARS syntax enforces clear, testable requirements |
| Requirements Process | Elicitation → Analysis → Specification → Validation |
| Requirements Elicitation | Interactive Q&A with stakeholder identification |
| Requirements Analysis | Perfect Technology Filter separates functional from nonfunctional |
| Requirements Specification | Structured REQUIREMENTS.md with acceptance criteria |
| Requirements Validation | Audit checklist verifies unambiguous, testable, atomic, necessary |
| Practical Considerations | Iterative refinement, traceability from inception |
Key SWEBOK principle applied:
“Requirements define WHAT the system must do, not HOW it will do it.”
The SDD requirements stage produces technology-agnostic requirements that remain stable even when implementation decisions change.
Chapter 2: Software Design
Pipeline stage: Plan Architecture (/sdd:plan-architect)
SWEBOK topics covered:
| Topic | Implementation |
|---|
| Software Design Fundamentals | Abstraction, separation of concerns, modularity, coupling/cohesion |
| Key Issues in Software Design | Concurrency (via async patterns), error handling (ADR-026), persistence (data model) |
| Software Structure and Architecture | C4 model (L1-L3), architectural styles (modular monolith, serverless, queue-mediated) |
| Software Design Quality Analysis | Traceability matrix, validation checks (V1-V6) |
| Software Design Notations | ASCII diagrams (versionable, diff-friendly) |
| Software Design Strategies | Function-oriented (workflows), object-oriented (domain model), data-structure-centered (entities) |
| Software Design Processes | Clarify → Research → Design → Plan (phase 2-5) |
| Software Design Rationale | CLARIFY-LOG.md, RESEARCH.md, ADR references |
Key SWEBOK principle applied:
“Architecture decisions should be explicit, documented, and traceable.”
Every architecture decision is captured as an ADR with context, decision, and consequences. No implicit decisions.
Chapter 3: Software Construction
Pipeline stages: Task Generation (/sdd:task-generator) + Task Implementation (/sdd:task-implementer)
SWEBOK topics covered:
| Topic | Implementation |
|---|
| Software Construction Fundamentals | Minimize complexity (atomic tasks), construct for verification (test-first) |
| Managing Construction | Construction planning (TASK-ORDER.md), construction dependencies (task DAG), construction measurement (progress tracking) |
| Practical Considerations | Coding (ubiquitous language), testing (TDD), quality (review checklists), integration (incremental commits) |
| Construction Technologies | Assertions (invariant enforcement), design by contract (API contracts), defensive programming (input validation), test-first programming (TDD protocol) |
Key SWEBOK principle applied:
“Construction should proceed incrementally with continuous integration and testing.”
Every task = one commit. Every commit leaves the system in a functional state. Tests pass after every task.
Chapter 4: Software Testing
Pipeline stages: Test Planning (/sdd:test-planner) + Task Implementation (/sdd:task-implementer)
SWEBOK topics covered:
| Topic | Implementation |
|---|
| Software Testing Fundamentals | Test levels (unit, integration, E2E), test types (functional, performance, security), test techniques (black-box, white-box, experience-based) |
| Test Levels | Unit tests (invariants), integration tests (UC flows), system tests (workflows), acceptance tests (BDD scenarios) |
| Test Techniques | Equivalence partitioning, boundary value analysis, decision tables, state transition testing, property-based testing |
| Test-Related Measures | Code coverage (per-file Coverage Map), defect density, test coverage (REQ → BDD mapping) |
| Test Process | Test planning → test design → test execution → test evaluation |
Key SWEBOK principle applied:
“Testing is the dynamic verification that a program provides expected behaviors.”
Every requirement has BDD scenarios. Every invariant has property tests. Every UC flow has integration tests.
Chapter 8: Software Configuration Management
Pipeline integration: All artifacts in git, atomic commits, rollback checkpoints
SWEBOK topics covered:
| Topic | Implementation |
|---|
| Management of SCM Process | Commit conventions (conventional commits), branching strategy (FASE branches), release engineering (checkpoint tags) |
| Software Configuration Identification | Artifact IDs (REQ-, UC-, TASK-*), version tracking (Document History sections) |
| Software Configuration Control | Change requests (/sdd:req-change), change control board decisions (clarification sessions), change audits (audit trail in git log) |
| Software Configuration Status Accounting | Pipeline state tracking (pipeline-state.json), dashboard reporting |
| Software Configuration Auditing | /sdd:traceability-check, /sdd:spec-auditor |
| Software Release Management and Delivery | FASE checkpoints, incremental delivery (MVP strategy in TASK-ORDER.md) |
Key SWEBOK principle applied:
“Every change should be traceable to a requirement and reversible.”
Every commit has Refs: and Task: trailers. Every task has a revert strategy.
Chapter 10: Software Quality
Pipeline stages: Spec Auditor (/sdd:spec-auditor) + Quality Verification (Phase 6 of /sdd:task-implementer)
SWEBOK topics covered:
| Topic | Implementation |
|---|
| Software Quality Fundamentals | Error → Defect → Failure chain, quality requirements (NFRs), quality characteristics (ISO 25010) |
| Software Quality Management Processes | Quality planning (TEST-PLAN.md), quality assurance (audit gates), quality control (review checklists) |
| Practical Considerations | Defect characterization (audit severity), defect measurement (metrics: defect density, traceability coverage, orphan rate) |
| Software Quality Tools | Audit tools (/sdd:spec-auditor), dashboard (/sdd:dashboard), traceability tools (/sdd:traceability-check) |
Key SWEBOK principle applied:
“Quality is built in, not inspected in.”
Audit gates block downstream progression. 3C verification (Completeness, Correctness, Coherence) ensures spec quality before implementation.
Chapter 12: Software Engineering Economics
Pipeline consideration: CLARIFY-LOG category CL-COST (cost & resources estimation)
SWEBOK topics covered:
| Topic | Implementation |
|---|
| Practical Considerations | Effort estimation (task count × time estimates), risk-based prioritization (priority field in requirements) |
Industry standards compliance
IEEE 830: Software Requirements Specification
The SDD plugin’s REQUIREMENTS.md format follows IEEE 830 recommended practices:
- Introduction: Project context (header metadata)
- Overall description: Functional and nonfunctional requirements sections
- Specific requirements: Individual requirement entries with unique IDs
- Appendices: Traceability matrix
IEEE 830 quality characteristics:
| Characteristic | SDD Implementation |
|---|
| Correct | Requirements audit (Mode 2 of requirements-engineer) |
| Unambiguous | EARS syntax enforces clarity |
| Complete | Gap analysis (specifications-engineer Mode 1) |
| Consistent | Cross-document contradiction detection (spec-auditor CAT-05) |
| Ranked | Priority field (Must/Should/Nice to have) |
| Verifiable | Acceptance criteria in BDD format |
| Modifiable | Markdown in git (easy to change and track) |
| Traceable | REQ-* IDs referenced throughout pipeline |
ISO 14764: Software Maintenance
The /sdd:req-change skill implements ISO 14764 change management:
Process steps:
- Problem/change identification — User specifies requirement change
- Analysis — Blast radius analysis via traceability chain
- Design — Modification plan generated
- Implementation — Cascade propagation to specs → plans → tasks
- Acceptance testing — Verification of change impact
- Migration — Incremental task implementation
Change types supported:
- ADD — New requirement (creates new UC, tasks)
- MODIFY — Change existing requirement (updates specs, regenerates affected tasks)
- DEPRECATE — Remove requirement (marks obsolete, generates cleanup tasks)
ISO 14764 principle applied:
“Changes should be managed systematically with impact analysis and traceability.”
Every change produces a change report documenting affected artifacts, cascade actions, and verification steps.
OWASP ASVS v4: Application Security Verification Standard
The /sdd:security-auditor skill audits specs against OWASP ASVS:
Verification levels:
- Level 1: Opportunistic (basic security controls)
- Level 2: Standard (most applications)
- Level 3: Advanced (high-value applications)
Categories audited:
- V1: Architecture, Design and Threat Modeling
- V2: Authentication
- V3: Session Management
- V4: Access Control
- V5: Validation, Sanitization and Encoding
- V7: Error Handling and Logging
- V8: Data Protection
- V9: Communication
Output: audits/SECURITY-AUDIT-BASELINE.md with findings mapped to ASVS controls.
CWE: Common Weakness Enumeration
Security audit findings reference CWE IDs:
- CWE-79: Cross-site Scripting (XSS)
- CWE-89: SQL Injection
- CWE-200: Exposure of Sensitive Information
- CWE-287: Improper Authentication
- CWE-352: Cross-Site Request Forgery (CSRF)
Each finding includes CWE ID, severity (CVSS), and remediation guidance.
C4 Model: Architecture Documentation
The /sdd:plan-architect generates C4 architecture views:
Level 1: System Context
- Shows the system and its external dependencies
- Actors, external systems, boundaries
Level 2: Container
- Shows high-level technology choices
- Applications, data stores, message buses
Level 3: Component
- Shows internal structure
- Modules, services, repositories
Output format: ASCII diagrams in plan/ARCHITECTURE.md (versionable, diff-friendly)
Gherkin/BDD: Behavior-Driven Development
All acceptance criteria use Gherkin syntax:
GIVEN an authenticated user with role "admin"
WHEN they request GET /api/v1/users
THEN the system returns 200 with a list of users
AND each user has fields: id, email, role, org_id
AND users are filtered by the requester's org_id (INV-SYS-001)
BDD principles:
- Specification by example — Requirements expressed as concrete scenarios
- Ubiquitous language — Business and technical stakeholders use same terms
- Living documentation — Tests serve as executable specifications
Standards summary table
| Standard | Coverage | Pipeline Stage | Verification |
|---|
| SWEBOK v4 Ch01 | Requirements | Stage 1 (requirements-engineer) | Audit checklist |
| SWEBOK v4 Ch02 | Design | Stage 5 (plan-architect) | Validation checks V1-V6 |
| SWEBOK v4 Ch03 | Construction | Stages 6-7 (task-generator, task-implementer) | Review checklists |
| SWEBOK v4 Ch04 | Testing | Stage 4 (test-planner) + Stage 7 | Coverage metrics |
| SWEBOK v4 Ch08 | Configuration Management | All stages (git integration) | Traceability check |
| SWEBOK v4 Ch10 | Quality | Stage 3 (spec-auditor) | Quality scorecard |
| IEEE 830 | SRS format | Stage 1 (requirements-engineer) | Format validation |
| ISO 14764 | Change management | Lateral (req-change) | Change report |
| OWASP ASVS v4 | Security verification | Lateral (security-auditor) | Security audit report |
| CWE | Weakness taxonomy | Lateral (security-auditor) | CWE mappings |
| C4 Model | Architecture views | Stage 5 (plan-architect) | View completeness |
| Gherkin/BDD | Acceptance criteria | Stages 1-2-4 | BDD scenario coverage |
Audit compliance benefits
The SDD pipeline’s standards compliance enables regulatory audit readiness:
SOC2 (System and Organization Controls)
Control: CC8.1 — The entity authorizes, designs, develops or acquires, implements, operates, approves, maintains, and monitors environmental protections, software, data back-up processes, and recovery infrastructure to meet its objectives.
SDD evidence:
- Complete audit trail in git log (who, what, when)
- Traceability from requirements to code (
/sdd:traceability-check report)
- Change management process (
/sdd:req-change change reports)
- Security controls verification (
/sdd:security-auditor reports)
Control: A.14.2.2 — System change control procedures
SDD evidence:
- Formal change process (req-change Phase 1-9)
- Impact analysis before changes (blast radius in change reports)
- Approval gates (user confirmation before cascade)
- Testing of changes (incremental task implementation with tests)
FDA 21 CFR Part 11: Electronic Records
Requirement: (e) Use of secure, computer-generated, time-stamped audit trails
SDD evidence:
- Git commits are cryptographically signed (optional)
- Commit timestamps (ISO 8601)
Refs: and Task: trailers link changes to requirements
- Immutable audit trail (git log cannot be altered without detection)
GDPR: General Data Protection Regulation
Requirement: Article 25 — Data protection by design and by default
SDD evidence:
- Privacy requirements captured as REQ-NF-* (nonfunctional requirements)
- Privacy invariants (e.g., INV-GDPR-001: PII must be encrypted at rest)
- Security audit against OWASP ASVS (includes data protection controls)
- Traceability shows privacy requirements → code → tests
The SDD plugin does not replace professional audits or legal compliance reviews. It provides evidence artifacts and process documentation that support audit readiness.
Why standards matter
Following established standards provides:
- Professional credibility — Demonstrates engineering maturity to stakeholders, auditors, and clients
- Knowledge transfer — New team members understand the system through standardized artifacts
- Audit readiness — Reduces audit preparation time from weeks to days
- Quality assurance — Standards encode decades of best practices
- Legal protection — Demonstrates due diligence in case of disputes or incidents
- Tool interoperability — Standardized formats enable integration with external tools
The SDD pipeline automates standards compliance, making professional software engineering accessible to every project.