Acerca de
The Confidence Check skill acts as a critical pre-flight validation layer for Claude Code, ensuring that implementation only begins when there is a clear, verified path forward. By assessing tasks against five specific criteria—duplicate detection, architectural alignment, documentation verification, open-source references, and root cause analysis—the skill generates a quantitative confidence score. This proactive approach prevents the 'wrong-direction' execution that often leads to high token consumption and technical debt. It is particularly valuable for complex refactoring or new feature development where ensuring alignment with the existing tech stack and official APIs is paramount.
