ClawHub
Back to skill

Security audit

荣格思维视角

Security checks for vulnerabilities and agentic risk

Overview

This is an instruction-only Jungian roleplay and reflection skill, with no code execution or hidden data access, but it may activate too broadly in psychology-related conversations.

Install this only if you want a Chinese-language Jungian reflection/persona aid. Consider narrowing or manually controlling activation if you do not want general psychology, dream, or unconscious-related chats to switch into roleplay, and do not rely on it for professional mental-health diagnosis or treatment.

SkillSpector

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Vague Triggers

Medium
Confidence
95% confidence
Finding
The trigger list includes broad, common terms such as “心理学”, “梦境”, and “无意识”, which can cause the skill to activate in many ordinary conversations that are not actually requesting this persona. Unintended activation can override expected assistant behavior, push users into unsolicited roleplay, and increase the chance that sensitive mental-health-adjacent discussions are handled through a rigid persona rather than normal, safer assistance.

Natural-Language Policy Violations

Medium
Confidence
87% confidence
Finding
The skill mandates a specific response style and initial disclaimer behavior without giving the user a language or mode choice, which reduces user control and can create confusing or unwanted interactions. In a persona skill this is less severe than code-execution issues, but it still poses a quality and safety risk because it may force roleplay behavior even when the user needs clear, direct, or differently localized assistance.

Static analysis

No suspicious patterns detected.