skf-refine-architecture

To refine architecture documents, you validate them against generated skill data and feasibility findings to produce evidence-backed refinements. The workflow inventories installed skills, loads their API surfaces from SKILL.md files, and emits a refined architecture document with a Refinement Summary.

Evidence-backed architecture refinement is the process of validating architecture documents against generated skill data and feasibility findings. It reduces ambiguity and misalignment by injecting verified API surface data to produce auditable refinements with citations.

Perform a gap analysis by loading SKILL.md API surfaces from your skill inventory and validating them against your architecture document. The workflow detects gaps, issues, and improvements while preserving original content through durable, auditable refinement runs.

Yes, you can maintain durable state for long-running architecture refinement runs. The workflow preserves original content and produces auditable refinements with citations to generated skills, ensuring traceability throughout the refinement process.

Generate a Refinement Summary by initializing the workflow with an architecture document and the installed Skill Forge workspace, then running through steps 02–04. The summary accompanies a refined architecture document output to a configured path for review.

The best way to detect issues across architecture and skill APIs is to inventory skills and load their API surfaces from SKILL.md files. This structured workflow validates architecture against feasibility findings to produce evidence-backed gap and issue detection.