A key ingredient of both mastery learning and competency-based education is giving students multiple opportunities to practice with rapid feedback such as that provided by autograders. The open-source PrairieLearn assessment authoring system makes it possible to create exceptionally rich interactive STAR assessments: skill-specific, tagged to course competencies, auto-graded, and randomizable, so that by generating multiple variants of a question, students can practice as much as they need (and in addition, exams can be randomized to thwart cheating and to allow “second chance” re-takes easily). PrairieLearn is now in regular use in over 800 courses at 20 universities; instructors representing these courses will showcase innovative online assessments they’ve created using PrairieLearn and which they’re willing to make available to the wider instructor community. Presenters will also be happy to discuss their experience authoring these assessments.