It is increasingly common for software and hardware systems to support touch-based interaction. While the technology to support this interaction is still evolving, common protocols for providing consistent communication between hardware and software are available. However, this is not true for gesture recognition -- the act of translating a series of strokes or touches into a system-recognizable event. Developers often end up writing code for this process from scratch due to the lack of higher-level frameworks for defining new gestures. Gesture recognition can contain a significant amount of work since it often involves complex, platform-specific algorithms. We present a domain-specific language that significantly simplifies the process of defining new gestures and allows them to be used across multiple hardware platforms.