I need a feature or autonomous agent that can oversee the development of an original idea or application, ensuring that the YOLO agent stays on track. This agent should enforce strict adherence to real implementations—eliminating placeholders, mock data, and hallucinated outputs. Additionally, it should proactively identify and correct errors while allowing flexibility to refine and optimize these guidelines as needed.