In the current computing continuum, edge computing is an important field of research. Edge computing is a paradigm that revolutionizes traditional cloud-centric computing models by decentralizing data processing and analysis closer to the data source, often at or near the network’s edge. This approach aims to alleviate the latency and bandwidth constraints associated with transmitting large volumes of data to distant cloud servers. By leveraging local computational resources, such as edge devices or servers, edge computing empowers real-time decision-making, enhances privacy, and enables applications in environments with limited or intermittent connectivity. Currently, there exist many different platforms for edge computing exist. However, these platforms often form a relatively encapsulated system, requiring a specific implementation or abstraction. Within this thesis, a proposal of an interoperable protocol for the integration of different edge platforms and devices is created. Different entities can use the protocol to provide or require computational power and resources, which allows efficient offloading in a heterogeneous environment. No homogenous platform is necessary, but heterogeneous devices can communicate via this protocol. Additionally, context is used to allow for specific optimization methods, e. g., to improve the connectivity or decrease the computation delay. Based on this protocol, a framework is implemented, called MetaEdge. This prototype is used to validate the suitability of the protocol and to show its effectiveness. The concept is proofed by creating multiple worker nodes which implement different edge platforms and runtimes. MetaEdge is able to orchestrate and coordinate these different tasks, and also use context of the worker and the network for different optimization strategies.