What problem does it solve? Deciding which supporting files (references, examples, prompts) each skill in a Claude Code plugin needs is often arbitrary, leading to bloated or under-resourced skills. This Skill provides a structured, score-based method to allocate MOAT (Materials, Ontology, Assets, Templates) assets per skill. ## Core Features & Use Cases - Complexity Scoring: Rates each skill across four dimensions (domain knowledge, procedure steps, edge cases, user interaction) to produce a LOW to VERY HIGH complexity score. - Depth Allocation: Maps complexity scores to MOAT depths (MINIMAL, STANDARD, FULL, DEEP) with override rules for spec-referencing or artifact-producing skills. - Asset Planning: Defines naming conventions, file purposes, and line-count estimates for references/, examples/, and prompts/ directories, then compiles everything into a moat-strategy.md document. - Use Case: After drafting an architecture plan for a new plugin, run this Skill to decide that a hook-validation skill needs a compatibility matrix reference and examples, while a simple utility skill needs only its SKILL.md. ## Quick Start Plan the MOAT strategy for my plugin using the architecture plan at docs/architecture-plan.md.
