ClawHub
Back to skill

Security audit

Patent Validator

Security checks across malware telemetry and agentic risk

Overview

This is a text-only patent research helper that drafts search strategies and does not request execution, credentials, persistence, or data access.

Before installing, understand that this skill can help organize patent-search queries and research questions, but it should not be treated as legal advice or as a determination that an idea is patentable, unique, or free from existing implementations.

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 (1)

Vague Triggers

Medium
Confidence
89% confidence
Finding
The activation examples include generic phrases such as asking for similar implementations or what to search for, which can plausibly occur outside a patent-specific workflow. In agent-routing systems that rely on loose semantic matching, this can cause the skill to activate on unrelated research or comparison requests and steer users into an unintended patent-oriented flow, potentially collecting sensitive concept details or producing irrelevant guidance.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.