Getting Started with the SUSE Multi-Linux Manager MCP Server and GitHub Copilot

Enterprise Linux environments are varied and that’s just a fact of life. In every data centre I’ve encountered, I’ve seen a blend of systems such as SUSE, Ubuntu, RHEL, and their various derivatives. This trend is also expanding into Azure subscriptions. AI assistants like GitHub Copilot have the capability to connect to these machines, execute commands, troubleshoot issues, and apply patches individually. However, if you’re overseeing hundreds or thousands of systems across different distributions, the real challenge isn’t whether AI can access your infrastructure. It’s about how well it can integrate with the central management tools where your inventory, patch orchestration, role-based access control (RBAC), and audit trails are managed.

SUSE has made a significant move to address this issue. Their Multi-Linux Manager MCP Server, which is built on the open-source Uyuni project, offers AI agents like GitHub Copilot a structured and secure way to interact with your existing management platform. It’s not about managing individual machines but about centralising your management processes. You won’t need to rip out the old systems or learn to navigate a new console. Instead, you’ll have a straightforward way to communicate with the management infrastructure you’ve already established.

This article will guide you through the functions of the MCP server, its significance within the Azure context, and how to integrate it with GitHub Copilot, so you can start using it right away.

The Model Context Protocol (MCP) acts as an open standard that outlines how AI models can connect with external tools and data sources. Imagine it as the USB-C of AI integrations; it provides a universal interface so that different clients (like GitHub Copilot, Claude Desktop, and Gemini CLI) can communicate with various servers (including Azure, SUSE, databases, and APIs) without needing custom code for every combination.

If you’re running Linux workloads in Azure—whether for SAP, High-Performance Computing (HPC), or traditional enterprise applications—the Multi-Linux Manager MCP server can offer a conversational interface for your infrastructure without the need to change your existing tools.

• Comprehensive Management Layer: Azure and Copilot already help you see your VMs across the fleet. The SUSE MCP server adds another layer, where you can manage patch schedules, track errata, conduct cross-distribution CVE audits, and oversee system groups—all from your Multi-Linux Manager instance.
• Unified Viewpoint: When linked with the Azure MCP Server, your AI assistant can seamlessly shift between Azure resource management and OS-level tasks in one conversation, handling various distributions supported by Multi-Linux Manager without needing to switch tools or contexts.

The MCP Server offers over 20 practical tools for managing your infrastructure on a daily basis. Rather than relying on a generic knowledge base, Copilot can enquire directly about your specific infrastructure.

• Inventory Access: You can retrieve a list of active systems across your fleet or pull detailed event histories for specific machines.
• Patch Management and CVE Handling: Copilot can swiftly audit your systems for pending updates or pinpoint machines that are vulnerable to any new CVEs.
• Operational Tasks: You can list system groups, register new systems, or schedule reboots as needed.

Allowing an AI agent to interact with production infrastructures raises concerns about precautionary measures. SUSE has taken a careful approach by designing the MCP server around a default “human-in-the-loop” security model.

• Read-Only by Default: The server comes configured with write actions disabled (UYUNI_MCP_WRITE_TOOLS_ENABLED=false).
• Explicit Approval: If you choose to enable writing capabilities, Copilot must ask for your explicit confirmation before carrying out state-changing actions such as applying patches or scheduling reboots.
• Enterprise-Level Authentication: The server supports OAuth 2.0, ensuring that the AI agent authenticates through your identity provider.
• Layered Governance: Combined with the Multi-Linux Manager’s role-based access control (RBAC), you’ll benefit from layered governance without needing to add extra approval systems.

AI-assisted operations that skip human oversight aren’t likely to be embraced by enterprises. Instead, operations that enhance human efficiency while keeping them firmly in control are the model likely to thrive.

This is the structure we’re working with:

1. SUSE Multi-Linux Manager – Executing on an Azure VM, controlling your Linux fleet across different distributions. This serves as the command centre for your systems, covering inventory, patching, and configuration. Available on the Azure Marketplace.
2. MCP Server – Operates as a container (Docker/Podman), either locally alongside your development environment or as a standalone HTTP service. The MCP Server container is accessible in the SUSE Registry and is supported by a safe and reliable software supply chain.
3. GitHub Copilot – Utilised within VS Code or the command line. It’s configured to access the MCP Server as a tool source, allowing it to send natural language requests and receive structured responses from your infrastructure.
4. Your Linux fleet on Azure – Whatever the Multi-Linux Manager is overseeing for you. The MCP server is distribution-agnostic, which is exactly the idea behind Multi-Linux Manager.

• A running instance of SUSE Multi-Linux Manager that manages your Linux estate.
• Docker or Podman installed on your workstation (for local deployment) or access to a remote MCP server instance.
• GitHub Copilot with agent mode enabled (in VS Code or command line).

For local deployment, retrieve the container and set it up to point at your Multi-Linux Manager instance, following the project documentation. For remote deployments, your admin can launch the server as a standalone HTTP service with OAuth 2.0 for validation.

In VS Code, open the Command Palette and type GitHub Copilot: Configure MCP Servers. Add your server to the configuration:

{
  “mcpServers”: {
“suse-multi-linux-manager”: {
  “type”: “http”,
  “url”: “https://your-mcp-server.example.com/mcp”
}
  }
}

Next, try a read-only query in GitHub Copilot:

“List all active systems managed by my SUSE Multi-Linux Manager.”

If your fleet’s inventory shows up, you’re successfully connected.

“Are any of my systems affected by CVE-2026-XXXX?”

“Show me all systems that require pending but unscheduled security patches.”

“Which systems are due for a reboot?”

The SUSE Multi-Linux Manager MCP server is open-source under the Apache 2.0 licence, based on the Uyuni project. The current version 0.5 is a technical preview. Any feedback should be sent to uyuni-project/uyuni#10562, and bugs reported in GitHub Issues.

The question wasn’t whether AI could access your Linux operations but whether it could effectively work through the management tools that drive decisions at scale. SUSE has built the connection to this layer, while GitHub Copilot serves as the conversational interface. Your fleet is already in place—now it’s time to connect.