Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

Gateway Restart Profiler

v0.1.2

OpenClaw Gateway 重启性能分析工具。分析 Gateway 重启速度、排查启动慢的原因、生成性能报告、诊断启动问题。支持 Windows (PowerShell) 和 Linux (Bash) 平台。触发场景包括:分析 gateway 重启、检查启动时间、gateway 太慢了、生成性能报告、诊断...

0· 163·0 current·0 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for tstj007/gateway-restart-profiler.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "Gateway Restart Profiler" (tstj007/gateway-restart-profiler) from ClawHub.
Skill page: https://clawhub.ai/tstj007/gateway-restart-profiler
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install gateway-restart-profiler

ClawHub CLI

Package manager switcher

npx clawhub@latest install gateway-restart-profiler
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Suspicious
medium confidence
Purpose & Capability
The skill claims to profile Gateway restarts and the included Linux script does exactly that (stop/start gateway, tail logs, compute phase timings, output HTML/text reports). However, the SKILL.md references a Windows PowerShell script (gateway-profile.ps1) that is not present in the file manifest, and the instructions require running the 'openclaw' CLI (openclaw gateway stop/start) although required binaries lists none — the skill omits declaring a real dependency on the openclaw binary.
Instruction Scope
Instructions direct the agent to stop and start the Gateway and parse local log files under ~/.openclaw/logs or /tmp, which is within the stated purpose. They also explicitly request the agent run the commands elevated=true (admin/root). The script reads only local logs and writes reports to ~/.openclaw/logs and /tmp; it does not contain obvious external data exfiltration, but it does load Chart.js from a CDN for the HTML report.
Install Mechanism
No install spec or external downloads are used; the skill is instruction-only plus a shell script. This minimizes installation risk because no third-party archive is fetched or executed at install time.
Credentials
The skill declares no required environment variables or credentials, which is appropriate. The script only accesses local log paths and writes reports under the user's home/temp directories; no secrets or unrelated credentials are requested.
!
Persistence & Privilege
The SKILL.md instructs the agent to execute the profiling scripts with elevated privileges (elevated=true) and to stop/start the Gateway. Running with elevation is a significant privilege and will interrupt service; while this may be necessary to manage the Gateway, it increases risk and should be accepted explicitly by the user and run in a controlled window. The skill does not request persistent installation (always: false).
What to consider before installing
Before installing or running this skill: 1) Understand that it will stop and start your Gateway and requires elevated (admin/root) execution — schedule downtime or run in a test environment. 2) Verify the presence and origin of the 'openclaw' CLI on the host (the script calls 'openclaw gateway stop/start' but the skill does not declare this dependency). 3) Note the SKILL.md mentions a Windows PowerShell script that is not included; do not assume Windows support is present. 4) Review the included shell script yourself (it is present) to confirm it does only the expected log parsing and report generation; it does not appear to exfiltrate data, but it does load Chart.js from a CDN in the HTML report. 5) If you require stricter isolation, run the script manually in a safe environment (or on a staging Gateway) rather than granting an agent autonomous elevated execution. If you want me to, I can point out exact lines in the shell script that perform sensitive actions or suggest a safer invocation approach.

Like a lobster shell, security has layers — review code before you run it.

latestvk97fv8778gzfz89dawaa8kct2d85cthr
163downloads
0stars
3versions
Updated 4d ago
v0.1.2
MIT-0

Gateway Restart Profiler

分析 OpenClaw Gateway 重启时的各阶段耗时,生成性能报告并给出优化建议。

支持平台

  • Windows: PowerShell 5.0+
  • Linux: Bash 4.0+, 需安装 bc 计算器

使用方法

Windows

powershell -ExecutionPolicy Bypass -File <skill>/scripts/gateway-profile.ps1

Linux

bash <skill>/scripts/gateway-profile.sh

代理调用方式

当用户要求分析 Gateway 重启性能时,OpenClaw 代理应使用以下方式调用:

Windows:

exec(command="powershell -ExecutionPolicy Bypass -File <skill>/scripts/gateway-profile.ps1", elevated=true)

Linux:

exec(command="bash <skill>/scripts/gateway-profile.sh", elevated=true)

其中 <skill> 应替换为实际路径,例如:

  • Linux: /home/tstj/.openclaw/workspace/skills/gateway-restart-profiler
  • Windows: C:\Users\<user>\.openclaw\workspace\skills\gateway-restart-profiler

代理调用时会自动重启 Gateway(先 stop 再 start),监听日志解析各阶段耗时,生成 HTML 可视化报告和文本摘要。

输出内容

  1. 实时日志监控 - 高亮显示各启动阶段
  2. HTML 可视化报告 - Chart.js 图表(柱状图 + 饼图)
  3. 详细数据表 - 含耗时百分比和可视化进度条
  4. 优化建议 - 自动识别超过30秒的阶段并给出建议
  5. 文本摘要 - 保存到 temp 目录

分析的阶段

阶段说明
配置加载读取 openclaw.json
身份验证连接认证服务
HTTP服务启动 Web 服务
CanvasCanvas 挂载
MCP服务MCP 协议服务
心跳服务心跳监控
模型加载加载 AI 模型
频道启动启动 QQ/Telegram 等频道
插件系统加载插件

优化建议解读

阶段耗时可能原因
配置加载 > 30s配置文件过大或损坏
身份验证 > 30s网络问题或认证服务故障
模型加载 > 30sOllama 响应慢,建议换用 API 模式
频道启动 > 30sQQ/Telegram 连接异常
插件系统 > 30snpm 依赖缺失,需重新安装

日志文件位置

  • Windows: C:\Users\<user>\AppData\Local\Temp\openclaw\
  • Linux: ~/.openclaw/logs/

示例输出

HTML 报告(用浏览器打开)包含:

  • 📊 耗时分布饼图 - 各阶段占总耗时比例
  • 📈 横向柱状图 - 按耗时排序,一目了然
  • 📋 详细数据表 - 毫秒/秒/占比/可视化进度条
  • 💡 优化建议 - 超过30秒的阶段高亮提示

控制台同时输出:

阶段耗时明细 (按耗时降序):
  频道启动           45000ms (45s)   ########################
  模型加载           32000ms (32s)   ################
  配置加载           12000ms (12s)   #########

优化建议:
  - [频道启动] 检查QQ/Telegram连接是否正常
  - [模型加载] Ollama响应慢,考虑换用API模式

[HTML图表报告已保存: ~/.openclaw/logs/gateway-profile-2026-04-12.html]
  → 用浏览器打开 HTML 文件查看可视化图表!

Comments

Loading comments...