Context-aware systems are capable of perceiving the physical environment where they are deployed and adapt their behavior, depending on the available information and how it is processed. Ambient Intelligence (AmI) represents context-aware environments that react and respond to the requirements of people. While different models can be used to implement adaptive context-aware systems, BDI multi-agent systems are especially suitable for that, due to their belief-based reasoning. Different BDI architectures, however, use different reasoning processes, therefore providing different adaptability levels. In each architecture , contextual information is adherent to a specific belief structure, and the context-related capabilities may vary. We propose a framework that can be used by BDI agents in a multi-architecture scenario in order to modularly acquire context-aware capabilities, such as learning, additional reasoning abilities, and interoperability. When this framework is combined with an existing BDI agent, the result is an augmented agent.