LLM-powered AI agent systems are bringing new perspectives to enterprise design, operations, and computing (EDOC), particularly in environments where agents can act autonomously and independently, referred to as agentic systems. These capabilities unlock new opportunities for automation, enabling agents to perform intellectually demanding tasks that were previously reserved for humans, while still allowing human oversight of key decisions. However, a critical challenge remains: ensuring clear accountability within these systems across humans and AI Agents, which may involve complex chains of authorization and delegation. As individual agents act independently, pinpointing responsibility becomes increasingly difficult. This paper proposes a novel solution to this problem: a domain-specific language (DSL) based on the ISO ODP Enterprise Language standard, precisely defining the roles and interactions between actors in the enterprise landscape. The DSL is implemented using textX, a contemporary tool-chain which provides rapid prototyping ecosystem. The aim is to provide user-friendly syntax while following the precise semantics of ODP enterprise language.