ClawHub
Back to skill

Security audit

ADC专利监控（ADC Patent Monitoring）

Security checks for vulnerabilities and agentic risk

Overview

This appears to be a legitimate ADC patent reporting skill, with incomplete package metadata and a third-party chart script that users should notice before use.

Before installing, confirm you have the intended PatSnap/Eureka MCP access and are comfortable sending patent queries through that service. Also note that generated HTML reports load Chart.js from a public CDN unless the template is changed to use a local copy.

SkillSpector

Vulnerability Patterns
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Description-Behavior Mismatch

Medium
Confidence
88% confidence
Finding
The manifest metadata does not match the declared skill purpose: it names a different skill and provides no description, which can mislead users, reviewers, or downstream tooling about what the package actually does. In security-sensitive agent ecosystems, this kind of identity mismatch weakens trust boundaries, impairs review, and can help unauthorized or unreviewed functionality hide behind inaccurate labeling.

Static analysis

No suspicious patterns detected.