Model Context Protocol evolves as Nutanix integrates AI agents into cloud operations

Nutanix has released an open-source Model Context Protocol (MCP) server for its Nutanix Cloud Platform (NCP). This integration allows AI assistants and developer tools, such as GitHub Copilot and Claude Code, to connect directly to cloud infrastructure via the Prism v4 API. The system is designed to enable AI agents to perform routine infrastructure tasks—such as checking system health or triggering automation—using natural language instructions while operating within established security and governance frameworks.

To manage the risks of direct AI access to production systems, the MCP server utilizes fine-grained role-based access control, API throttling, and auditing to track agent-initiated commands. This ensures that AI agents operate within defined permissions rather than having unrestricted access.

Concurrently, the Model Context Protocol itself is evolving toward a stateless protocol model. This shift in the MCP specification changes how sessions are handled, moving away from protocol-level sessions toward a stateless architecture. This transition impacts how developers scale MCP servers and manage Streamable HTTP, facilitating more efficient connections between AI applications and external tools, resources, and data.