Genetically modified mice are used as models for a variety of human behavioral conditions. However, behavioral phenotyping can be a major bottleneck in mouse genetics because many of the classic protocols are too long and/or are vulnerable to unaccountable sources of variance, leading to inconsistent results between centers. We developed a home-cage approach using a Chora feeder that is controlled by-and sends data to-software. In this approach, mice are tested in the standard cages in which they are held for husbandry, which removes confounding variables such as the stress induced by out-of-cage testing. This system increases the throughput of data gathering from individual animals and facilitates data mining by offering new opportunities for multimodal data comparisons. In this protocol, we use a simple work-for-food testing strategy as an example application, but the approach can be adapted for other experiments looking at, e.g., attention, decision-making or memory. The spontaneous behavioral activity of mice in performing the behavioral task can be monitored 24 h a day for several days, providing an integrated assessment of the circadian profiles of different behaviors. We developed a Python-based open-source analytical platform (Phenopy) that is accessible to scientists with no programming background and can be used to design and control such experiments, as well as to collect and share data. This approach is suitable for large-scale studies involving multiple laboratories.