What problem does it solve? Codebases accumulate shallow, tightly-coupled modules whose interfaces are nearly as complex as their implementations, making them hard to test, navigate, and refactor safely. This Skill explores a codebase to surface that architectural friction and turns it into actionable, well-designed refactoring proposals. ## Core Features & Use Cases - Friction-driven exploration: Uses an Explore sub-agent to navigate the codebase organically, flagging clusters where understanding one concept requires bouncing between many small files. - Parallel interface design: Spawns 3+ sub-agents that each produce a radically different interface for the deepened module (minimal, flexible, common-case-optimized, ports-and-adapters), then compares them with an opinionated recommendation. - Dependency classification and testing strategy: Categorizes dependencies as in-process, local-substitutable, ports-and-adapters, or external mocks, and prescribes boundary tests that replace shallow unit tests. - GitHub issue generation: Creates a refactor RFC via gh issue create using a structured template covering problem, proposed interface, dependency strategy, and testing plan. - Use Case: A team with a microservices backend notices integration bugs hiding in the seams between small service modules. Run this Skill to get a ranked list of deepening candidates, pick one, review competing interface designs, and receive a ready-to-discuss GitHub issue. ## Quick Start Ask the assistant to explore this repository and propose module-deepening refactoring opportunities as GitHub issues.
