llm-judge
LLM-as-judge methodology for comparing code implementations across repositories. Scores implementations on functionality, security, test quality, overenginee...
