Nutanix launches MCP server for AI cloud operations

Nutanix has launched an MCP server for Nutanix Cloud Platform, aimed at customers using AI tools to automate cloud operations.

The product implements the Model Context Protocol specification and connects AI agents and developer tools to Nutanix Cloud Platform through the Nutanix Prism v4 API. It allows IT teams to use plain-language prompts to trigger infrastructure actions while keeping existing access controls in place.

The launch reflects a broader push by technology suppliers to embed generative and agent-based AI tools in routine infrastructure management. Nutanix is targeting customers running hybrid and multicloud environments, where day-to-day administration often requires specialist knowledge and tightly controlled permissions.

The MCP server acts as a pass-through layer between AI assistants and the cloud platform. Nutanix cited GitHub Copilot, Claude Code and Cursor as examples of assistants that can connect to the system.

According to Nutanix, the software translates everyday language requests into API actions on cloud infrastructure. The design is intended to reduce the risk of unauthorised or poorly scoped changes by enforcing corporate security policies at the point of execution.

Control layer

The system is built on the Nutanix Prism v4 API Gateway, which handles execution, governance and security controls. As a result, connected AI tools inherit the same fine-grained access controls and governance model already used in the platform.

Operators can use AI assistants to assess system health, gather diagnostics and prepare automated workflows while retaining human oversight. Nutanix described this as a human-in-the-loop approach intended to keep final control with IT staff rather than autonomous agents.

The governance model includes role-based access control, throttling and metering, audit logging, and support for long-running asynchronous tasks. These functions are designed to limit agents to explicitly authorised API calls, regulate traffic, record which agent initiated a command, and let agents track the progress of larger jobs.

That focus on governance addresses a central issue in enterprise AI adoption. Businesses may be willing to use AI assistants to speed up repetitive administration, but many remain cautious about letting software agents interact directly with live production systems without clear controls and accountability.

Developer focus

Nutanix is also pitching the server to developers building internal automation tools. AI coding assistants can use system definitions provided through the server to generate scripts in languages including Python, Go, Java, JavaScript and PowerShell, as well as curl and other REST- and JSON-compatible formats.

The aim is to give teams a way to create custom agents or workflow tools for Nutanix infrastructure without building each connection from scratch. Users can also configure security permissions through a single interface.

Open-source licensing may widen the product's appeal among developers and IT teams that want to test AI-driven automation without being tied to a proprietary connector model. Nutanix said customers can use the server with their own AI agents.

For Nutanix, the release adds another layer to its broader effort to position its platform as a control point for infrastructure spread across private and public cloud environments. The company has long focused on simplifying infrastructure management, and the addition of an MCP server places that effort within the current market for agentic AI tools.

In a statement accompanying the launch, Thomas Cornely, executive vice president of product management at Nutanix, said: "Nutanix has a long history of using automation to simplify complex cloud operations. The MCP server is an exciting next step on our journey, allowing us to bring the power of autonomous AI agents to cloud management."

Cornely also stressed the control framework around the new software: "By creating a secure gateway between AI tools and our platform, we are giving customers the confidence to safely use AI to operate and govern their hybrid multicloud environments."