概要
The Verification Skill ensures high-reliability outputs by mandating a strict 'evidence-before-claims' protocol for Claude Code. It eliminates uncertainty and false success reports by requiring specific commands, fresh execution results, and binary assertions before any task transition or completion claim. By banning speculative language like 'should work' and requiring explicit command output, it creates a robust audit trail for development activities, ensuring that every fix or feature is objectively proven to be functional.
