What problem does it solve? Routines that do many things at once are hard to name, test, understand, and modify. This Skill provides diagnostic signals and extraction techniques to split unfocused routines into single-responsibility functions. ## Core Features & Use Cases - Diagnostic Signals: Detect unfocused routines using the "and" test, parameter counts over 7, lengths over 200 lines, vague names, and mixed abstraction levels. - Extraction Techniques: Extract by responsibility, by abstraction level, and by complex conditional logic into well-named boolean functions. - Cohesion Framework: Classify routines across seven cohesion levels from coincidental to functional, targeting functional cohesion. - Use Case: When reviewing a 300-line order handler that validates, prices, checks inventory, persists, and emails, split it into a thin orchestrator calling focused routines like validate_order_request and calculate_pricing. ## Quick Start Review this function and extract each responsibility into its own focused routine with a clear single-purpose name.