Self-feeding robot arm is an assistive device for elderly people or patients in hospitals which lack assistance from nurse or doctor. However, some simple tasks which may not be suitable for a meal with many types of food like soup, noodle, or meat. For solve this problem to have a meal with full of nutrients and create diversity for it, in this research, a toolkit was developed that can be used in many foods base on operation of multifunction CNC machines. By using image processing and machine learning, the camera can identify which kind of food in the bowl then choose the suitable tools. Moreover, it can self-regulate actions for that type of food. After training and testing, the system provides a fast response time and high accuracy.