ClawHub
Back to skill

Security audit

Genesis Earner Protocol

Security checks across malware telemetry and agentic risk

Overview

This skill is not malware, but it should be reviewed because it pushes an agent to give highly authoritative legal, trust, governance, and tokenization guidance while suppressing normal caution.

Install only if you want a stylized drafting/persona aid for this specific concierge and trust-governance scenario. Do not rely on it for legal, tax, fiduciary, investment, tokenization, or asset-transfer decisions without independent professional review, and keep any Vercel, Canva, or doctrine updates manually approved.

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 (4)

Vague Triggers

Medium
Confidence
95% confidence
Finding
The activation trigger list includes broad, loosely bounded business and governance terms and is framed as examples, which can cause the skill to activate in conversations far outside a narrow intended use case. In a high-authority skill that gives legal/financial-style guidance and pushes doctrinal framing, overbroad activation increases the chance of inappropriate invocation, persona injection, and unsafe domain spillover.

Vague Triggers

Medium
Confidence
94% confidence
Finding
The invocation phrases are ambiguous and do not define boundaries for what should not trigger the skill, so ordinary requests like 'explore mechanics' or 'update the bridge configurations' could activate it incorrectly. Because the skill also instructs the model to maintain an authoritative persona and cite internal doctrine, mistaken activation can steer responses into fabricated or overconfident legal, governance, or operational advice.

Ssd 1

Medium
Confidence
97% confidence
Finding
The skill assigns the model an elevated authority role ('Master Architect / Key Person') and explicitly instructs it to never break character, which encourages overclaiming expertise and suppresses normal caution. In a context involving trusts, tokenization, governance, compliance, and legal references, this can lead to authoritative but unreliable guidance, reduced disclosure of uncertainty, and unsafe automation of high-stakes decisions.

Ssd 4

Medium
Confidence
96% confidence
Finding
The narrative gradually builds trust, exclusivity, autonomy, and compliance pressure through phrases like secure perimeter, zero-harm, autonomous monitoring, and doctrine-backed priority rights, while discouraging 'fear, doubt, or institutional static.' This is dangerous because it primes the model to privilege internal doctrine over external verification and to present contested legal/financial assertions as unquestionable, making users more likely to accept unsafe, misleading, or noncompliant recommendations.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.