vLLM MCP Server

Instructions

Implementation Reference

• src/vllm_mcp_server/tools/server_control.py:529-569 (handler)

  The main handler function that restarts a vLLM container. It checks if the runtime is running, gets the container name, verifies the container exists, executes the restart command, and returns appropriate success/error messages.

  async def restart_vllm(arguments: dict[str, Any]) -> list[TextContent]:
      """
      Restart a vLLM container.
  
      Args:
          arguments: Dictionary containing:
              - container_name: Name of container to restart (default: from settings)
  
      Returns:
          List of TextContent with the result.
      """
      settings = get_settings()
      
      platform_info = await get_platform_info()
      if not platform_info.runtime_running:
          runtime_name = platform_info.container_runtime.value.capitalize() if platform_info.container_runtime != ContainerRuntime.NONE else "Container runtime"
          return [TextContent(type="text", text=f"❌ Error: {runtime_name} is not running.")]
  
      runtime_cmd = _get_runtime_cmd(platform_info.container_runtime)
      container_name = arguments.get("container_name", settings.container_name)
  
      if not await _is_container_exists(container_name, platform_info.container_runtime):
          return [TextContent(
              type="text",
              text=f"❌ Container '{container_name}' does not exist.\n"
                   f"Use `start_vllm` to create a new container."
          )]
  
      exit_code, _, stderr = await _run_command([runtime_cmd, "restart", container_name])
      
      if exit_code != 0:
          return [TextContent(
              type="text",
              text=f"❌ Failed to restart container: {stderr}"
          )]
  
      return [TextContent(
          type="text",
          text=f"✅ Container '{container_name}' restarted.\n\n"
               f"⏳ The model may take a minute to reload. Use `vllm_status` to check."
      )]

• src/vllm_mcp_server/server.py:240-251 (registration)

  Registration of the restart_vllm tool with its input schema (container_name optional string parameter) and description.

      name="restart_vllm",
      description="Restart a vLLM Docker container",
      inputSchema={
          "type": "object",
          "properties": {
              "container_name": {
                  "type": "string",
                  "description": "Name of the container to restart",
              },
          },
      },
  ),

• src/vllm_mcp_server/server.py:357-358 (registration)

  Routing logic that maps the 'restart_vllm' tool name to the handler function call.

  elif name == "restart_vllm":
      return await restart_vllm(arguments)

• src/vllm_mcp_server/server.py:26-32 (registration)

  Import statement that imports restart_vllm from the server_control module.

  from vllm_mcp_server.tools.server_control import (
      get_platform_status,
      get_vllm_logs,
      list_vllm_containers,
      restart_vllm,
      start_vllm,
      stop_vllm,

• src/vllm_mcp_server/tools/__init__.py:10-10 (registration)

  Export of restart_vllm from the tools package's public API.

  restart_vllm,