skf-analyze-source

Monorepo decomposition involves mapping public export surfaces and cross-unit integration points to identify tightly coupled areas. By detecting service and module integration boundaries, you can separate discrete units and generate actionable skill briefs calibrated to your analysis depth tier.

By applying tier-aware boundary detection to a multi-service codebase, the system identifies discrete skillable units and maps their dependencies. It then automatically produces conformant skill-brief.yaml files validated against the Agent Skills schema for immediate downstream use.

Yes, boundary detection supports tier-aware analysis depth configurations. Whether using Quick, Forge, Forge+, or Deep tiers, the system calibrates its scanning of service, package, and module boundaries to produce appropriately detailed skill briefs and export maps.

The best way to automate AI agent skill authoring is to scan existing codebases for discrete skillable units and their integration boundaries. This approach automatically produces schema-conformant skill-brief.yaml files and export maps, eliminating manual and inconsistent unit identification.

Yes, you can map cross-unit integration points and public export surfaces to flag tightly coupled areas. This mapping identifies dependencies between discrete units and highlights stack skill candidates during the codebase analysis process.