improve-codebase-architecture

To improve codebase architecture, explore the repository to surface shallow modules and seams that constrain testability and AI navigability, returning a prioritized, actionable list of proposed changes with involved files and current friction.

Architectural seams are shallow modules or tightly-coupled points in a codebase that constrain testability and modularity by making the codebase less AI-navigable and harder to refactor without regressions.

Refactor tightly-coupled modules by framing deepening candidates using CONTEXT.md vocabulary and ADR decisions, ensuring domain alignment and traceability by documenting decisions when candidates contradict existing ADRs.

Yes, you can inspect a repository to identify deepening opportunities and receive a prioritized list of candidates detailing involved files, current friction, plain-English solutions, and expected locality and leverage.

Prioritize refactoring tasks by evaluating actionable deepening candidates that explicitly outline expected locality and leverage, allowing you to target tightly-coupled modules that maximize architectural improvement and AI navigability.

Document ADR conflicts during a refactor when identified deepening candidates contradict existing ADR decisions, ensuring that architectural changes maintain domain alignment and traceability within the codebase.