Skip to content
mcp/harbor
Grok

NVSentinel Lab for Grok

io.github.omerfsen/nvidia-nvsentinel-mcp

Test NVIDIA NVSentinel on a kind lab: read its state and inject faults into a simulated GPU.

client:Grok transport:stdio runtime:oci

Install NVSentinel Lab in Grok

.mcp.json (in your project root)

Held in this page only — never stored, logged, or sent anywhere but back to your screen. 

{
  "mcpServers": {
    "nvidia-nvsentinel-mcp": {
      "type": "stdio",
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "docker.io/omerfsen/nvidia-nvsentinel-mcp:0.5.1"
      ],
      "env": {
        "NVSENTINEL_MCP_KUBECONFIG": "<NVSENTINEL_MCP_KUBECONFIG>",
        "NVSENTINEL_MCP_ALLOW_CHANGES": "<NVSENTINEL_MCP_ALLOW_CHANGES>",
        "NVSENTINEL_MCP_LAB_CONTEXT": "<NVSENTINEL_MCP_LAB_CONTEXT>"
      }
    }
  }
}

Grok Build reads Claude Code's .mcp.json and Claude Desktop's config unchanged, so a server you have already set up there needs nothing new. The Grok web app and X integration do not take custom servers. Grok docs

NVSentinel Lab in other clients

NVSentinel Lab MCP server