What problem does it solve? Teams and AI assistants routinely produce abstractions, layers, theories, and design patterns that look elegant but add no explanatory or predictive value over the concrete code they replace. This Skill provides a single decision rule to detect over-abstraction and over-engineering before they calcify into the codebase or knowledge base. ## Core Features & Use Cases - Justification Test: For each abstraction, identify the concrete thing it replaces, then check whether it explains more, predicts more, or compresses (fewer assumptions, same predictions). - Three-Way Verdict: Classify each audited item as keep (earns its place), cut (pure restatement or decoration), or downgrade to an explicitly labeled unverified bet locked as a draft. - Dual-Layer Application: The same rule applies to conceptual/theoretical abstractions and to code-level over-design, complementing code review and simplification passes. - Use Case: While reviewing a design document proposing a new architectural layer, run the audit and receive a per-item verdict such as "cut — restates the existing module in heavier vocabulary" with a one-line reason. ## Quick Start Ask the AI to run the over-justify audit on the abstraction or design draft you just proposed and return a keep, cut, or bet verdict with a one-line reason for each item.
