Use when auditing a codebase for semantic duplication – functions that perform the same task but have different names or implementations. This is especially useful for LLM-generated codebases where new functions are often created instead of reusing existing ones.