ClawHub
Back to skill

Security audit

36计

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed strategy-analysis skill, but it needs review because it can apply deceptive or manipulative tactics to broad everyday strategy requests without clear ethical limits.

Review before installing. This skill does not appear to run code or access your data, but it may steer broad strategy questions toward adversarial, deceptive, or manipulative recommendations. Use it only where that framing is appropriate, and prefer explicit invocation such as /stratagem rather than broad auto-triggering.

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 (4)

Vague Triggers

Medium
Confidence
84% confidence
Finding
The README advertises 'Smart Triggering' that auto-activates on unspecified strategic keywords, but it does not define boundaries, precedence, or consent requirements. In an agent environment, broad and vague trigger rules can cause the skill to engage on unrelated user inputs, leading to unintended strategic advice, context capture, or instruction interference in conversations that were not meant to invoke this skill.

Vague Triggers

Medium
Confidence
95% confidence
Finding
The activation criteria are broad enough to trigger on ordinary requests about strategy or 'how should I deal with,' which can cause the skill to intercept conversations outside its intended niche. That increases the chance of unsolicited strategic framing and delivery of manipulative or adversarial advice in benign contexts, especially because the skill is user-invocable and optimized to auto-route into deeper analysis.

Vague Triggers

Medium
Confidence
88% confidence
Finding
The routing table uses broad natural-language triggers such as '怎么办', 'how should', 'strategy for', '分析/analyze', and competitive keywords that commonly appear in ordinary queries. In this skill context, that can cause overbroad activation or escalation into deeper strategic analysis for users who did not intend to invoke this skill, increasing the chance of inappropriate handling, misleading advice, or unnecessary collection/processing of sensitive business-competition context.

Ssd 4

Medium
Confidence
90% confidence
Finding
The framework operationalizes covert influence, asymmetry creation, psychological exploitation, and phased control as recommended output structure, not merely historical analysis. In practice this can be used to generate manipulative business, negotiation, or interpersonal tactics that facilitate deception, coercive influence, or unfair competitive conduct, and the rich workflow makes such misuse easier rather than incidental.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.