ClawHub

Xiong Xiaoge Perspective V2

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only roleplay skill with no code or data access; its main risk is confusing persona-style business commentary with real investment advice or the real person’s views.

Install only if you want a roleplay-style business perspective based on public information. Use explicit invocation, keep the disclaimer visible, and do not rely on it for specific investment decisions, stock picks, private views, or current factual claims without independent verification.

SkillSpector

By NVIDIA
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
87% confidence
Finding
The trigger list includes broad conceptual phrases such as "先锋精神", "早期投资", "全球化视野", and "长期陪伴" that can easily appear in ordinary conversations unrelated to this specific persona. This can cause unintended activation of the skill, leading the assistant to switch into a roleplaying investment-advisor style without clear user intent, which may confuse users and alter the system’s response behavior unexpectedly.

Natural-Language Policy Violations

Medium
Confidence
84% confidence
Finding
The skill instructs the agent to respond directly as the real person in first person and to avoid breaking character unless the user explicitly exits. Without explicit user consent or locale/language negotiation, this can create deceptive impersonation and force a Chinese-language/persona response even when the user did not clearly request that mode, reducing transparency and increasing the chance of misleading advice attribution.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal