Context-awareness is crucial for developing cyber-physical systems to enable dynamic behavior. Context-oriented programming (COP) aims to improve the definition of context-dependent behavioral variations. However, existing layer-based COP approaches have two major limitations: layered methods are specified only for single contextual dimensions and restrict adaptations to method replacement or input/output filtering, limiting expressiveness for complex context dependencies. This paper presents an approach to context-oriented programming supporting multiple contextual dimensions. Behavior definition is supported by a composition system that generates context-dependent methods enabling fine-grained code insertion at arbitrary locations within base methods. The approach is implemented as an embedded domain-specific language in Julia for the Contexts.jl library. Evaluation demonstrates improved scalability and flexibility compared to traditional layer-based approaches, enabling sophisticated context-dependent adaptations. This advancement facilitates the development of adaptive applications in domains requiring multi-dimensional contexts, such as autonomous vehicles and smart infrastructure.