What problem does it solve?
Turning a requirements analysis into an implementable system design is error-prone: modules get rewritten unnecessarily, interface contracts are retrofitted, and design decisions lose traceability to requirements. This Skill guides an AI agent through a disciplined four-stage pipeline that produces a Requirements Design Specification grounded in the actual codebase.
Core Features & Use Cases
- Codebase-Grounded Design Exploration: Analyzes the existing project code and architecture before designing, with effort-tiered subagent exploration (Low/Medium/High).
- Architecture & Interface Change Planning: Produces module add/modify/protect/freeze plans, Mermaid architecture change diagrams, and interface contracts with compatibility impact.
- Template-Driven Document Generation: Assembles SDD-style design documents from modular components via Node.js scripts, with dynamic checklists for quality review.
- Feasibility Verification & Review: Delegates red-team feasibility checks and deliverable reviews to subagents using dedicated checklists.
- Use Case: After completing requirements analysis for a new feature, run this Skill to generate a design specification covering module changes, data models, interfaces, DFx strategy, and SR-AR decomposition, ready for development planning.
Quick Start
Use the aet-req-design skill to turn my requirements analysis specification into a system design document based on the current project codebase.