ClawHub
Back to skill

Security audit

match-technology-transfer-opportunities-ip

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed technology-transfer analysis and report-writing workflow with scoped external lookups and user-approved file output.

Before installing, be prepared to approve any external patent or market-source lookup and any HTML output path explicitly. Do not provide confidential technology-transfer materials unless you intend the agent to analyze them, and treat the generated report as decision support requiring human legal, commercial, and compliance review.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
88% confidence
Finding
The skill instructs use of external patent connectors, authorized sources, and report-writing scripts, which implies network access and filesystem reads/writes despite no declared permissions. This mismatch is dangerous because it obscures the actual execution capabilities of the skill, weakening least-privilege review and increasing the chance of unauthorized data egress or file modification, especially given the skill explicitly handles potentially confidential technology-transfer materials.

Description-Behavior Mismatch

Medium
Confidence
94% confidence
Finding
The module is functionally a procurement-record normalizer and buyer-candidate extractor, which does not align with the stated skill purpose of technology-transfer partner matching. In this context, that mismatch is dangerous because it can silently drive downstream recommendations from irrelevant public-procurement evidence, producing materially misleading partner identification and decision-support reports while appearing legitimate and auditable.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.