derive-logical-architecture

What problem does it solve?

This skill resolves the hard architectural ambiguity of turning selected product capabilities into a layered logical structure that is traceable end-to-end, cycle-safe, and grounded in an existing systems inventory.

Core Features & Use Cases

• Selects components from a systems inventory: builds architecture by choosing systems/sub-systems instead of inventing new components.
• Applies a pinned or KB-derived layer model: places every selected component into the locked layer model for that Stage 3 checkpoint.
• Wires intent-correct edges with sync modes: records whether relationships are sync, async, or hybrid as design intent.
• Validates architectural integrity: blocks synchronous-only cycles and ensures every selected capability has at least one end-to-end path from the entry layer to a serving component.
• Surfaces decisions for auditability: emits a decision manifest that captures the rationale, grounding sources, and alternatives for downstream orchestrator actions.

Quick Start

Call the derive-logical-architecture skill during arch Stage 3 to produce logical-architecture.yaml and a decision manifest that trace every selected capability to a serving component.