Computer-Aided Assessment has a history nearly as long as computing itself. The earliest documented reference to using computers in an attempt to simplify grading dates back to 1959 at Rochester Polytechnic Institute, where a computer program was used to test the behavior of student's machine-language submissions [17]. Work in the area has continued in an ad-hoc fashion over the ensuing 45 years, and can give us a general idea of the components that are regarded by the computer-science education community as necessary for a CAA platform to be useful. To date there is still a notable lack of a definitive tool for CAA that meets all of the technical requirements, has sufficient developer backing, and is intuitive, powerful, and usable enough to gain widespread acceptance. Sampling the developments that have been made and the projects that have been attempted over the last 45 years can give us a good map of exactly what is needed. In fact, several recent CAA papers specifically enumerate the design goals that their designers were striving for. It is upon the shoulders of these developments that Agar stands, and much of the discussion on the design of Agar in Chapter 4 is informed by these papers.