DME MCP Server
Exposes DME storage O&M actions as MCP V1 tools, enabling management of storage modules such as SAN, NAS, and storage via natural language. It provides per-module MCP endpoints and a root server to enumerate available tools.
README
DME MCP Server
MCP Server for DME storage O&M
Exposes 16 action modules (427 actions) from dme-python-sdk as MCP V1 tools
- MCP V1 (
mcp>=1.28,<2, FastMCP) - Endpoints: one shared endpoint, one path per module, format
<endpoint>/mcp/v1/<module>(e.g.http://127.0.0.1:8000/mcp/v1/san) - Root MCP Server:
/mcp/v1exposes alist_toolstool that enumerates every module endpoint and its tools (see Root MCP Server) - Tool naming: dot-separated
<topic>.<action_key>(e.g.san.lun.list) - Annotation: every tool carries
meta = {"topic", "subtopic", "action"} - Docstring parsing: action function docstrings are parsed into the tool's
description/inputSchema(parameter descriptions) /outputSchema(Returns fields); supports both the Chinese (default branch) and English (main-enbranch) docstrings ofdme-python-sdk - Blacklist: modeled after
cli.py;~/.config/pydme/blacklist.jsontakes precedence. Every blacklisted tool gets an extra optionalaccept_riskparameter (default from--accept-risk/DME_ACCEPT_RISK); calling it withoutaccept_risk=trueis rejected with aRISK_BLOCKEDresponse telling the caller to set it explicitly
Installation
# Install dependency (Chinese)
pip install git+https://github.com/agentic-data-ops/dme-python-sdk.git
# Or install dependency (English)
pip install git+https://github.com/agentic-data-ops/dme-python-sdk.git@main-en
# Install the MCP Server
pip install git+https://github.com/agentic-data-ops/dme-mcp-server.git
Usage
# streamable-http (default); DME connection params match the pydme CLI.
# Listens on 127.0.0.1:8000 unless --mcp-server or DME_MCP_SERVER is set.
dme-mcp-server --mcp-server 0.0.0.0:8000 \
--endpoint https://192.168.1.100:26335 --user admin --password xxx
# Or use environment variables (omitting --mcp-server uses the 127.0.0.1:8000 default)
export DME_API_ENDPOINT=https://192.168.1.100:26335
export DME_API_USERNAME=admin
export DME_API_PASSWORD=xxx
dme-mcp-server
# stdio mode (--mcp-server is just a startup switch)
dme-mcp-server --mcp-server x --mcp-transport stdio \
--endpoint ... --user ... --password ...
Connecting
Each module is an independent MCP server. Point your MCP client at one of these URLs:
- Per module:
http://<host>:<port>/mcp/v1/<module>— e.g.http://127.0.0.1:8000/mcp/v1/san,/mcp/v1/nas,/mcp/v1/storage - Root:
http://<host>:<port>/mcp/v1— root server, exposes only thelist_toolstool
Root MCP Server
Call the list_tools tool on the root server to enumerate every module endpoint and its tools:
tools/call {"name": "list_tools", "arguments": {}}
Response (JSON text):
[{"mcp_server_path": "http://<host>:<port>/mcp/v1/<module>",
"tools": [{"name": "<tool_name>", "description": "<tool_description>"}]}]
Use the returned mcp_server_path values as the client URL to connect to the corresponding module server.
Options
| Option | Environment variable | Description |
|---|---|---|
--endpoint/-e |
DME_API_ENDPOINT |
DME API address |
--user/-u |
DME_API_USERNAME |
Username |
--password/-p |
DME_API_PASSWORD |
Password |
--token |
DME_API_AUTH_TOKEN |
Auth token (optional; skips login when provided) |
--timeout |
— | Request timeout (default 90) |
--no-cache-auth-token |
— | Disable token caching |
--accept-risk |
DME_ACCEPT_RISK |
Default accept_risk=true for blacklisted high-risk tools (clients can still override per call) |
--mcp-server |
DME_MCP_SERVER |
Listen address host:port (default: 127.0.0.1:8000) |
--mcp-transport |
— | streamable-http (default) / stdio |
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。