ClawHub
Back to skill

Security audit

mine-patentable-inventions-ip

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed patent invention-mining workflow that uses patent research tools and includes approval, confidentiality, and legal-boundary safeguards.

Before using this skill, confirm what confidential invention details may be shared with PatSnap or other research sources, use anonymous labels when needed, and have patent counsel verify legal, jurisdictional, infringement, standards, and translation-sensitive conclusions.

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
79% confidence
Finding
The skill description is extremely broad and can plausibly trigger on many normal R&D, IP, competitor, standards, or innovation-analysis requests. Over-broad routing increases the chance the agent invokes this skill in contexts involving confidential technical disclosures or legal-adjacent analysis without sufficiently specific user intent, causing unnecessary data exposure and workflow misapplication.

Natural-Language Policy Violations

Medium
Confidence
86% confidence
Finding
Forcing English interface and output without user opt-in can cause silent translation of technical or legal concepts, increasing the risk of meaning drift in patent, standards, and disclosure-sensitive workflows. In this skill’s context, nuance matters: mistranslation of claim terms, dates, materials, or disclosure constraints can produce incorrect search scope, legal framing, or inadvertent confidentiality mistakes.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.