Skip to main content
Glama
ygh11231
by ygh11231

AutoDL MCP Server

管理 AutoDL GPU 云实例的 MCP Server,支持实例生命周期管理、SSH 远程操作、文件传输和 GPU 监控。

功能

实例管理

  • autodl_list_instances — 列出所有弹性部署

  • autodl_get_instance — 获取实例详情

  • autodl_stop_instance — 停止实例(释放 GPU)

  • autodl_start_instance — 启动实例(分配 GPU)

  • autodl_delete_instance — 删除实例

  • autodl_set_replicas — 设置副本数(0=停止,>0=启动)

容器管理

  • autodl_list_containers — 列出部署下的容器

  • autodl_stop_container — 停止特定容器

  • autodl_get_container_events — 查看容器事件日志

GPU 库存与镜像

  • autodl_get_gpu_stock — 查询区域 GPU 库存

  • autodl_list_images — 列出私有镜像

  • autodl_get_regions — 列出可用区域

SSH 远程操作

  • autodl_ssh_connect — SSH 连接远程实例

  • autodl_ssh_disconnect — 断开 SSH 连接

  • autodl_run_command — 远程执行命令

  • autodl_upload_file — 上传文件(SCP)

  • autodl_download_file — 下载文件(SCP)

  • autodl_check_gpu — 查看 GPU 状态(nvidia-smi)

  • autodl_check_logs — 查看日志文件

  • autodl_list_ssh_connections — 列出活跃 SSH 连接

Related MCP server: projtool

安装

1. 克隆仓库

git clone https://github.com/ygh11231/autodl-mcp-server.git
cd autodl-mcp-server

2. 安装依赖

pip install -r requirements.txt

3. 设置 API Token

AutoDL 控制台 → 设置 → 开发者 Token 获取。

export AUTODL_API_TOKEN="your_token_here"

4. 测试运行

python server.py

正常启动后不会有输出,MCP 协议通过 stdin/stdout 通信。

在 Hermes 中配置

编辑 ~/.hermes/config.yaml,添加:

mcp_servers:
  autodl:
    command: "/path/to/your/python"
    args: ["/path/to/autodl-mcp-server/server.py"]
    env:
      AUTODL_API_TOKEN: "your_token_here"

重启 Hermes 后,所有工具会自动注册为 mcp_autodl_*

在 Claude Code 中配置

编辑 ~/.claude/.mcp.json

{
  "mcpServers": {
    "autodl": {
      "type": "stdio",
      "command": "/path/to/venv/bin/python",
      "args": ["/path/to/autodl-mcp-server/server.py"],
      "env": {
        "AUTODL_API_TOKEN": "your_token_here"
      }
    }
  }
}

使用示例

> 列出我的所有实例
> 停止实例 xxxx-xxxx-xxxx
> 启动实例 xxxx-xxxx-xxxx,设置 2 个副本
> 查询西北企业区的 GPU 库存
> SSH 连接我的实例,查看 GPU 状态
> 把我的代码上传到远程实例

项目结构

autodl-mcp-server/
├── server.py          # MCP 服务器入口
├── autodl_api.py      # AutoDL REST API 客户端
├── ssh_manager.py     # SSH 连接管理器
├── requirements.txt   # Python 依赖
├── README.md          # 本文件
└── LICENSE            # MIT 许可证

许可证

MIT

Maintenance

ActivityStale
ResponsivenessSyncing

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Connectors

Related MCP Servers

  • A
    license
    Not graded
    quality
    F
    maintenance
    Enables comprehensive management of Vast.ai GPU cloud instances, including searching for GPU offers, creating and managing instances, executing remote commands via SSH, and monitoring background tasks for ML training workflows.
    11
    MIT
  • A
    license
    Not graded
    quality
    B
    maintenance
    Enables ML researchers to manage experiments across local and remote AutoDL GPU instances, including experiment creation, training launch, run polling, and report writing via Claude Code.
    1
    MIT
  • A
    license
    Not graded
    quality
    C
    maintenance
    Enables management of AutoDL GPU resources including instance lifecycle, balance queries, elastic deployments, and notifications through typed MCP tools.
    2
    MIT
  • A
    license
    Not graded
    quality
    B
    maintenance
    Enables AI agents and MCP clients to manage AutoDL GPU instances programmatically, including creating, listing, powering on/off, executing commands, transferring files, and enforcing automatic shutdown.
    MIT

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/ygh11231/autodl-mcp-server'

If you have feedback or need assistance with the MCP directory API, please join our Discord server