概要
This skill implements a robust framework for authoring and maintaining Claude Code skills through a process called 'TDD for documentation.' It ensures that every new capability is rigorously tested against baseline agent behaviors, optimized for search discovery (CSO), and written for maximum token efficiency. By following the Red-Green-Refactor cycle—writing test scenarios, watching agents fail, and then drafting the skill—it eliminates workflow shortcuts and guarantees that agents follow full instructional sets rather than just surface-level descriptions.
