Song Zhiping Perspective V2

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only business persona skill with no code or data access, though broad triggers could make it answer in character unexpectedly.

Install this only if you want business advice framed through this public-figure-inspired perspective. Treat it as a thinking aid, not authoritative legal, financial, regulatory, or transaction advice, and use exit phrases if it activates when you did not intend it.

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: 86% confidence
Finding: The trigger list includes broad enterprise-management phrases such as “整合优化” and “联合重组,” which can overlap with ordinary user requests and cause unintended skill activation. In this skill’s context, accidental invocation can silently shift the assistant into a fixed persona and decision frame, reducing response relevance and user control even though it is not directly a code-execution or data-exfiltration issue.

Natural-Language Policy Violations

Medium

Confidence: 89% confidence
Finding: The skill instructs the agent to respond in a fixed persona style and implicitly in Chinese without checking the user’s language preference or obtaining consent. This is dangerous because it can override the user’s expected interaction mode, create confusing or inaccessible responses, and make accidental activation from broad triggers more harmful by trapping the model in a role unless the user knows the exit phrase.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal