AI导哈

SSH NAS管理工具

一个基于SSH的NAS远程管理服务，提供命令执行、文件操作、系统监控、Docker容器管理和服务状态检查等功能。

  • by lyzetam
  • Python
  • 本地部署
  • 更新于
在 GitHub 查看 lyzetam/mcp-ssh-nas

使用说明

mcp-ssh-nas

MCP server and LangChain tools for SSH-based NAS management -- execute commands, manage files, monitor Docker containers, and check system health.

Features

13 tools across 5 categories:

Category Tools Description
Core ssh_execute, ssh_status Execute arbitrary commands, check connection status
Files ssh_list_files, ssh_read_file, ssh_write_file, ssh_file_exists Full file operations over SSH
System ssh_system_info, ssh_disk_usage, ssh_memory_usage, ssh_process_list Hostname, OS, uptime, disk, memory, processes
Docker ssh_docker_ps, ssh_docker_logs List containers, view container logs
Services ssh_service_status Check systemd/service status

Installation

# Core library
pip install .

# With MCP server support
pip install ".[mcp]"

# With LangChain tools
pip install ".[langchain]"

# Everything
pip install ".[all]"

Configuration

Variable Required Default Description
NAS_HOST Yes (empty) NAS hostname or IP address
NAS_PORT No 22 SSH port
NAS_USER Yes (empty) SSH username
NAS_PASSWORD Yes (empty) SSH password

.env Example

NAS_HOST=192.168.1.100
NAS_PORT=22
NAS_USER=admin
NAS_PASSWORD=your_password_here

Quick Start

As MCP Server

mcp-ssh-nas

Add to Claude Code (~/.claude.json):

{
  "mcpServers": {
    "ssh-nas": {
      "type": "stdio",
      "command": "uv",
      "args": ["--directory", "/path/to/mcp-ssh-nas", "run", "mcp-ssh-nas"],
      "env": {
        "NAS_HOST": "192.168.1.100",
        "NAS_USER": "admin",
        "NAS_PASSWORD": "your_password"
      }
    }
  }
}

As LangChain Tools

from mcp_ssh_nas.langchain_tools import TOOLS, ssh_execute, ssh_docker_ps

# Use all 13 tools with an agent
agent = create_react_agent(llm, TOOLS)

# Or use individual tools
print(ssh_execute.invoke({"command": "uptime"}))
print(ssh_docker_ps.invoke({"all": False}))

As Python Library

from mcp_ssh_nas.client import SSHClient
from mcp_ssh_nas.operations import system, docker, files

client = SSHClient(host="192.168.1.100", user="admin", password="secret")
result = client.execute("ls -la /volume1/")
info = system.system_info(client)
containers = docker.docker_ps(client)

License

MIT

相关 MCP