ClawHub
Back to skill

Security audit

sunzibingfa

Security checks across malware telemetry and agentic risk

Overview

This is a markdown-only Sun Tzu advice skill with broad, ethically sensitive strategy guidance but no code execution, credential access, persistence, or hidden data behavior.

Install only if you want general strategy and interpersonal advice through a Sun Tzu lens. Do not rely on it for legal, medical, mental health, financial, workplace discipline, crisis, or highly sensitive relationship situations, and review outputs for manipulative, coercive, or overly competitive framing.

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 (3)

Vague Triggers

High
Confidence
95% confidence
Finding
The activation scope is extremely broad, covering generic 'how to' questions, dilemmas, decision analysis, interpersonal advice, and Sun Tzu topics. In an agent system, this can cause the skill to trigger on a large fraction of ordinary conversations, overriding more appropriate skills and steering users toward manipulation- or competition-framed guidance in contexts where it may be unsuitable.

Vague Triggers

Medium
Confidence
88% confidence
Finding
The activation rule says to activate for questions related to interpersonal conduct, methods, competition, leadership, or decision-making, but does not define boundaries for when not to activate. This ambiguity increases unintended invocation risk, making behavior unpredictable and potentially displacing safer or more specialized guidance.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The document extensively operationalizes deception, pressure, inducement, and interpersonal manipulation tactics for workplace, social, and competitive contexts without clear ethical boundaries, misuse warnings, or harm-minimization guidance. In an agent skill that may be activated for real-life advice, this increases the risk that users receive normalized manipulative strategies that can damage trust, enable coercive behavior, or escalate conflicts.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.