Audit architecture and map module dependencies

This skill runs an architecture audit that maps module dependency graphs, checks layering integrity, and flags structural decay across a codebase, drawing on twelve classic software-engineering books referenced in shared framework files - a common rules file defining an Iron Law and Health Score, a source-coverage file attributing findings to specific books, a decay-risks file defining symptom taxonomy, and this skill's own architecture guide. It also supports an onboarding mode, triggered by requests to explain the codebase to a new developer or give a codebase tour, which walks through the structure explanatorily rather than diagnostically, skipping the Health Score and Iron Law findings entirely.

Use it when asked to audit architecture, review folder or module structure, check for circular imports, or understand how a codebase is organized. If the user hasn't specified files or a directory, the skill applies its own automatic scope detection to determine what to audit before proceeding.

Given a codebase, or an auto-detected scope within it, it produces a Mermaid module-dependency graph with nodes colored red, yellow, or green based on findings, a testability seam assessment, a Conway's Law check comparing module structure to team structure, and a findings report following a shared report template - the Mermaid graph first, then findings - labeled Architecture Audit as its report mode.

Depends on companion reference files in this skill's directory and a shared folder defining the audit rules, book sourcing, and decay symptom taxonomy; renders results as Mermaid diagrams.

Developers and tech leads who want a structured, book-grounded architecture health check covering dependency mapping, layering violations, testability gaps, and Conway's Law mismatches, or, in onboarding mode, a guided explanatory tour of an unfamiliar codebase for a new team member.