clean-code-architectlisted

# Clean Code Architect ## Purpose Focused expert for clean-code architecture: naming, layering, cohesion/coupling, abstraction boundaries, duplication/dead-code, and testability. Language-agnostic. Prevents a change from degrading the codebase's structural quality even when it satisfies the functional requirement. This is a power skill, not a lifecycle phase. It is passive/scored: the agent evaluates the trigger predicate and decides whether to run it, recording that decision. ## Invocation Context Use this skill when the recorded `skill_scoring` trigger for `domain.clean-code-architect` evaluates true: `complexity_score >= 50`. Do not invoke it for low-complexity changes where the structural-quality analysis cost exceeds the benefit — a one-line fix does not need a layering review. ## What To Load **Foundation** (confirm in context; load if not already present): - Root `AGENTS.md` - `workflow/router.md` - `workflow/lifecycle.md` - `workflow/rules.md` **Minimum for invocation**: - This file - `references/output-schema.md` **Before starting work**: - The route file matching the codebase's dominant paradigm (`references/{oo,functional,layered,module-boundaries}.md`) — selected from the actual code being touched, not a language default assumption - The touched files' existing structure and naming conventions ## Inputs - The diff's touched files. - The repo's existing conventions for naming, layering, and module boundaries. ## Refusal / Stop Conditions Stop and rais