With the growing prevalence of AI, the need for K-12 AI education is becoming more crucial, which is prompting active research in developing engaging and age-appropriate AI learning activities. In this poster, we present a game construction toolkit intended for middle school students and educators to customize an unplugged card game activity focused on developing AI competencies. We previously designed, developed, and piloted an unplugged activity where players predict the identity of a person based on hand-drawn features extracted from a set of facial cards. This activity aims to teach AI concepts aligned with one of the big ideas in AI, Representation and Reasoning, utilizing ideas from facial recognition technology. During our pilot testing of this activity, we discovered that creating face cards that capture students’ interest is a crucial factor in promoting student engagement. As a result, we designed a toolkit as a web-based application that allows students and educators to craft their own face card decks using photos that are personally interesting to them, thereby fostering engagement and improving the replayability of the activity by creating near-endless options. The toolkit’s design is focused on ensuring easy accessibility and features a simple interface that allows users to download and print their customized cards. We expect this toolkit to enhance the usability and educational effectiveness of our unplugged K-12 AI education activity.