ClawHub
Back to skill

Security audit

Cybernetic Thinking

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed thinking scaffold with a user-controlled installer, with install-documentation caveats users should handle carefully.

Install only from a source you trust, review the npx installer prompts before allowing links into multiple agents, and avoid copying the rm -rf or ExecutionPolicy Bypass commands blindly. If removing the old skill, verify the exact path first or rename it to a backup instead. Treat the retry/backoff sample as conceptual, not production-ready code.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Intent-Code Divergence

Medium
Confidence
96% confidence
Finding
The worked example is internally inconsistent and unsafe as instructional content: it references an undefined variable `attempt` and re-raises on the first exception, so it does not implement the documented retry/backoff loop. In a reasoning skill that teaches control strategies for production systems, broken sample code can mislead downstream agents or users into adopting faulty resilience logic, causing outages, retry storms, or failed recovery behavior.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
文档明确建议用户执行 `rm -rf ~/.config/kilo/skills/skill-systems-thinking` 删除本地目录，但没有提示这是破坏性操作、需要先确认路径，或建议先备份/列目录核对。虽然删除目标看起来是旧 skill 路径而不是任意系统路径，但在 AI 助手/自动化场景中，用户或代理可能机械照抄执行，导致配置或本地定制内容丢失。

Missing User Warnings

Medium
Confidence
95% confidence
Finding
README 建议使用 `powershell -ExecutionPolicy Bypass -File init.ps1`，这会绕过 PowerShell 的执行策略保护，却没有附带风险说明或更安全替代方案。即便脚本本身可能无害，要求用户关闭一层平台安全控制会降低对脚本来源、完整性和后续复制粘贴命令的防护，尤其在 AI 代理代执行时更危险。

Tool Parameter Abuse

High
Category
Tool Misuse
Content
本 skill 面向支持自定义 skill 的 AI 编程助手（Kilo、Claude Code、Aider、Cursor 等）。

> **0.4.0 改名**：包名从 `skill-systems-thinking` 改为 `skill-cybernetic-thinking`。装过旧名的用户请先删掉旧链接：`rm -rf ~/.config/kilo/skills/skill-systems-thinking`（路径按你的 agent 调整），再装新的。

### npx 安装（推荐）
Confidence
87% confidence
Finding
rm -rf ~

Tool Parameter Abuse

High
Category
Tool Misuse
Content
本 skill 面向支持自定义 skill 的 AI 编程助手（Kilo、Claude Code、Aider、Cursor 等）。

> **0.4.0 改名**：包名从 `skill-systems-thinking` 改为 `skill-cybernetic-thinking`。装过旧名的用户请先删掉旧链接：`rm -rf ~/.config/kilo/skills/skill-systems-thinking`（路径按你的 agent 调整），再装新的。

### npx 安装（推荐）
Confidence
87% confidence
Finding
rm -rf ~/.config/kilo/skills/

VirusTotal

60/60 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.