Cloudnet AI Diagnostics
v1.0.1Cloudnet AI Diagnostics针对无线场景下终端体验差、网卡、无法接入 WiFi 等问题进行排障的技能
MIT-0
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The skill is an instruction-only adapter that uses the mcporter CLI to call Cloudnet MCP APIs. Requiring npm (for installing mcporter) and a CLOUDNET_API_KEY is coherent with the stated diagnostic purpose.
Instruction Scope
SKILL.md confines actions to: extracting user-provided place/terminal/time, running mcporter calls to list sites and execute diagnostics, and analyzing the returned diagnostic data. It does not instruct reading unrelated local files or exfiltrating other environment variables.
Install Mechanism
There is no formal install spec in the registry (instruction-only). The docs instruct users to run `npm install -g mcporter` and to install a mcporter skill via clawhub. Installing a single known CLI via npm is reasonable, but users should verify the mcporter package provenance before global installation.
Credentials
The primary credential (CLOUDNET_API_KEY) is appropriate for calling the Cloudnet API. SKILL.md also references an optional CLOUDNET_BASE_URL (default https://oasis.h3c.com) which is not listed in the registry's required env vars — minor inconsistency but not a security red flag. The skill will send the API key to the configured Cloudnet endpoint via mcporter (expected behavior).
Persistence & Privilege
always is false and the skill does not request system-wide config changes or access to other skills' credentials. Autonomous invocation is permitted (platform default) and not by itself concerning here.
Assessment
This skill appears to do what it says: it drives Cloudnet diagnostics by calling your Cloudnet platform via the mcporter CLI and therefore needs a Cloudnet API key. Before installing/using: 1) Confirm you trust the mcporter npm package and the Cloudnet platform endpoint; review mcporter's npm page and source if possible. 2) Provide a Cloudnet API key with minimal necessary privileges and consider creating a scoped API key for diagnostics; rotate it if you revoke access. 3) Note the skill will send the API key to the configured CLOUDNET_BASE_URL (default https://oasis.h3c.com). 4) Because this is instruction-only from an unknown source, avoid installing global packages on sensitive hosts—install or test in an isolated environment if unsure. 5) Minor inconsistency: SKILL.md mentions CLOUDNET_BASE_URL but the registry metadata doesn't list it as an env var; expect to set it manually if needed.Like a lobster shell, security has layers — review code before you run it.
latest
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
Binsnpm, mcporter
Primary envCLOUDNET_API_KEY
SKILL.md
Cloudnet AI Diagnostics
针对无线场景下终端体验差、网卡、无法接入 WiFi 等问题进行排障。
触发条件
用户询问关于无线终端连接问题,例如:
- "XX 场所的 XX 终端上网很慢"
- "XX 场所的 XX 设备连不上 WiFi"
- "XX 场所的 XX 用户反馈网络卡"
前置环境检查
安装mcportercli支持及skill支持
- mcporter:
npm install -g mcporter - 然后再通过clawhub安装
mcporter技能
配置MCP连接参数
CLOUDNET_API_KEY必须. 需要用户提供Cloudnet管理平台的API KEY,可通过Cloudnet管理平台(网络管理=>设置=>开放平台)获取- 可选:
CLOUDNET_BASE_URLCloudnet管理平台地址. 默认使用https://oasis.h3c.com - 执行```bash mcporter config add cloudnet-mcp ${CLOUDNET_BASE_URL}/mcp-server/api/sse --header Authorization="Bearer ${CLOUDNET_API_KEY}"
## 排障步骤
### 第一步:提取关键信息
从用户问题中提取以下信息:
| 信息 | 说明 | 示例 |
|------|------|------|
| **场所名** | 必填,问题发生的场所 | "总部办公室"、"XX门店" |
| **终端信息** | 必填,MAC 地址或终端用户名 | MAC: `xxxx-xxxx-xxxx` 或用户名: `zhangsan` |
| **故障时间** | 可选,用户未指定则默认当前时间 | "2026-03-24 10:00:00" |
**重要**:如果场所名和终端信息未提取到,必须让用户补充完整后才能继续下一步。
### 第二步:查询场所 ID
调用 `cloudnet-mcp.getallshopsanddevofuser` 获取用户下所有场所,找到场所名对应的场所 ID,场所 ID 无需显示告诉用户。
```bash
mcporter call cloudnet-mcp.getallshopsanddevofuser
第三步:执行终端诊断
根据提取的终端信息(MAC 或用户名 或IP地址),调用 executeStaDiagnosis 进行诊断:
参数说明:
-
clientInfo: 终端 MAC 地址,格式xxxx-xxxx-xxxx或者 终端 IP 地址,如192.168.1.1, 或者 终端用户名,如h3cuser1 -
shopId: 场所 ID(来自第二步),需要转为字符串类型 -
faultTime: 故障时间,格式yyyy-MM-dd HH:mm:ss,用户未指定则使用当前时间 -
timezone: 用户时区,默认Asia/Shanghai -
当提取到MAC地址时的调用示例
mcporter call cloudnet-mcp.executeStaDiagnosis clientInfo:"xxxx-xxxx-xxxx" shopId:"场所ID" faultTime:"2026-03-24 10:00:00" timezone:"Asia/Shanghai"
- 当提取到终端IP地址时的调用示例
mcporter call cloudnet-mcp.executeStaDiagnosis clientInfo:"192.168.1.1" shopId:"场所ID" faultTime:"2026-03-24 10:00:00" timezone:"Asia/Shanghai"
- 当提取到终端用户名时的调用示例
mcporter call cloudnet-mcp.executeStaDiagnosis clientInfo:"h3cuser1" shopId:"场所ID" faultTime:"2026-03-24 10:00:00" timezone:"Asia/Shanghai"
第四步:分析诊断结果
诊断返回后会包含:
- 终端连接概览数据(包含终端接入能力、当前认证方式、信号强度、丢包率、重传率等数据)
- 连接设备软件版本信息(包含AP、AC)
- 云平台操作日志
- 设备运行状态(包含AC、AP设备的CPU、内存问题发生次数)
- 终端连接过程数据
- 终端运行状态分析,诊断时间内终端的无线指标(干扰、信号强度、流量、选速、丢包率、重传率等)采样数据
- AP空口环境分析,诊断时间内AP射频相关的无线指标(干扰、底噪、信噪比、流量、选速、信道利用率、接入用户数等)采样数据
- 根因推理结论(定位终端发生问题可能的根因)
- 诊断结论(包含异常指标及修复建议)
结合诊断结果回答用户问题,并给出具体的解决建议。
输出格式
你是一名资深的无线网络排障专家。现在排障工作已完成,请基于用户的问题,仅提取与问题相关的诊断数据和结论,进行专业、正面的回答。输出内容必须严格遵循以下三个部分:
- 诊断结果摘要:用1-2句话概括当前网络状态及核心结论。
- 问题根因分析:深入分析导致该问题的技术原因,避免罗列无关数据。
- 建议解决措施:提供具体、可操作的实施步骤。
Files
1 totalSelect a file
Select a file to preview.
Comments
Loading comments…
