Automated grading tools (autograders) enable instructors to quickly assess student programming submissions in a scalable fashion. Not only do they save staff time, they enable a virtuous cycle in which students can submit work, get immediate feedback, and make improvements before the final deadline, rather than finding out about shortcomings when it is too late to get credit for fixing them. This supports mastery-based and specifications grading.

Jacquard is a powerful open-source Java autograder for Gradescope. It is closely integrated with other code evaluation tools, including Checkstyle (code style adherence), PMD (static analysis), JUnit 5 (unit testing), and JaCoCo (code coverage and cyclomatic complexity). To encourage students to write good tests, Jacquard supports cross-testing, where student tests are graded on their ability to find bugs in hidden instructor code while not falsely reporting bugs in correct code. Jacquard also provides a powerful API for custom tests, such as that code uses language features in the manner required by the instructor.

Workshop attendees will learn how to use basic and advanced Jacquard features so they can apply Jacquard to their own assignments. Attendees will also have the opportunity to discuss how best to apply autograders in different contexts (such as homework assignments vs. tests).