Structured-Prompt-Driven Development (SPDD)

Summary

Structured-Prompt-Driven Development (SPDD) is an engineering method that treats LLM prompts as first-class, version-controlled artifacts to make AI-generated code governable, reviewable, and reusable. The workflow focuses on moving uncertainty "to the left" by using structured templates to align code generation with business requirements and engineering standards.

Key Points

• Utilizes the REASESS Canvas, a seven-part structure (Requirements, Entities, Approach, Structure, Operations, Norms, and Safeguards) to guide LLM generation from intent to execution.
• Implements a "closed-loop" workflow where requirements updates are applied to the prompt before the code, and code refactors are synchronized back to the prompt via /spdd-sync.
• Employs the openspdd command-line tool to automate the lifecycle, including domain analysis, canvas generation, and code production.
• Prompts act as the primary artifact for code reviews, replacing fragmented chat histories with a single, versioned source of truth.
• Enforces engineering governance through "Norms" (e.g., naming, observability) and "Safeguards" (e.g., security rules, performance limits) defined within the prompt structure.

Technical Details

The SPDD workflow is operationalized through the openspdd CLI, which provides specific commands for different stages of the development lifecycle. The process begins with /spdd-analysis to extract domain keywords and design direction, followed by /spdd-reasons-canvas to generate an executable blueprint. Code is then produced via /spdd-generate, which processes the canvas task-by-task based on defined "Operations." For validation, /spdd-api-test can generate cURL-based test scripts covering normal, boundary, and error scenarios.

The REASONS Canvas is architected into three functional layers:
1. Abstract (Intent & Design): Covers Requirements, Entities, Approach, and Structure to define the problem and system architecture.
2. Execution: Focuses on Operations, breaking the strategy into concrete, testable implementation steps.
3. Governance: Defines Norms (cross-cutting engineering standards) and Safeguards (non-negotiable boundaries like invariants and security rules).

Impact / Why It Matters

SPDD allows organizations to scale AI-assisted development from individual productivity gains to a controlled, team-wide capability. It ensures that as the volume of AI-generated code increases, the ability to maintain architectural alignment and code quality remains intact.