Security audit

guanguan-10agents

Security checks across malware telemetry and agentic risk

Overview

This is a Chinese-language HR advice skill with broad activation wording, but it does not contain code, hidden access, persistence, or destructive behavior.

Install this if you want Chinese-language HR drafting and people-management support. Treat labor-law, termination, compensation, and dispute guidance as general drafting help, verify it against the correct jurisdiction, and be aware the skill may activate on broad people-management wording.

SkillSpector

By NVIDIA

Vulnerability Patterns

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

Findings (2)

Vague Triggers

High

Confidence: 95% confidence
Finding: The trigger definition is extremely broad: it mandates activation for nearly any topic involving 'people' or management, with expansive keywords and an explicit instruction to trigger immediately. This can cause the skill to hijack unrelated conversations, override user intent, and route sensitive or ambiguous queries into HR-specific guidance without sufficient relevance checks.

Natural-Language Policy Violations

Medium

Confidence: 87% confidence
Finding: The skill description is written to force a Chinese-language/locale interaction model without any user opt-in or adaptive behavior. If auto-invoked, this can degrade usability, mis-handle user requests in other languages, and create incorrect assumptions about legal or labor-law context, which is especially risky for HR and employment guidance.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.