Complex skill mastery requires not only acquiring individual basic component skills, but also practicing integrating such basic skills. However, traditional approaches to knowledge modeling, such as Bayesian knowledge tracing, only trace knowledge of each decomposed basic component skill. This risks early assertion of mastery or ineffective remediation failing to address skill integration. We introduce a novel integration-level approach to model learners' knowledge and provide fine-grained diagnosis: a Bayesian network based on a new kind of knowledge graph with progressive integration skills. We assess the value of such a model from multifaceted aspects: performance prediction, parameter plausibility, expected instructional effectiveness, and real-world recommendation helpfulness. Our experiments based on a Java programming tutor show that proposed model significantly improves two popular multiple-skill knowledge tracing models on all these four aspects.