ClawHub
Back to skill

Security audit

litigation-risk-monitor

Security checks across malware telemetry and agentic risk

Overview

The skill fits its litigation-monitoring purpose, but it automatically sends sensitive business/legal queries to external services and generated reports load third-party JavaScript.

Review before installing for sensitive matters. Use it only if you are comfortable sending target assignee names and patent-litigation queries to PatSnap and web search, and avoid opening generated reports in restricted environments unless the CDN script is bundled locally and remote image fallbacks are disabled.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (8)

Lp3

Medium
Category
MCP Least Privilege
Confidence
93% confidence
Finding
The skill instructs the agent to perform file reads/writes and networked data collection, but it declares no corresponding permissions or capability boundaries. This creates a governance gap: an operator may assume the skill is passive documentation while it actually performs external retrieval and local report generation, increasing the chance of unintended data access or exfiltration.

Context-Inappropriate Capability

Medium
Confidence
90% confidence
Finding
The skill requires downloading external patent images and embedding them as base64 into the generated report, which expands the skill from patent-risk analysis into active retrieval and repackaging of remote content. This increases attack surface through untrusted external resources, can pull in unexpectedly large or malformed files, and may persist third-party content into local outputs without necessity for the core analytical task.

Context-Inappropriate Capability

Medium
Confidence
94% confidence
Finding
The generated HTML loads Chart.js from a public CDN, causing the report to execute third-party JavaScript whenever opened. That creates a supply-chain and privacy risk: a compromised CDN asset or network-path tampering could execute arbitrary code in the viewer's browser, and opening the report leaks access metadata to the CDN.

Description-Behavior Mismatch

Medium
Confidence
92% confidence
Finding
The renderer claims the report is offline/self-contained, but the generated HTML still fetches Chart.js from a public CDN and may load remote images via `abstract_image_url`. Opening the report can therefore leak sensitive patent/litigation viewing context, requester IP, and timing metadata to third parties, and it also breaks integrity assumptions because remotely hosted content can change or become unavailable.

Intent-Code Divergence

Medium
Confidence
90% confidence
Finding
The misleading 'inline/offline usable' comment masks actual network dependencies, which is security-relevant in this skill because reports concern litigation and patent intelligence that may be confidential. Users may open the HTML in restricted or sensitive environments expecting no egress, but the page can still contact external infrastructure for scripts and images, creating privacy and supply-chain risk.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill mandates fully automatic external searches and report generation without any user-facing warning or consent around privacy, data handling, or third-party transmission. Because inputs include applicant names, inventors, litigation details, and generated reports, the workflow can send potentially sensitive business intelligence to external services without informed user approval.

Missing User Warnings

Medium
Confidence
79% confidence
Finding
The skill explicitly performs automatic file generation without intermediate confirmation, which can cause unreviewed write operations and creation of potentially sensitive reports from externally sourced data. In a system handling legal and patent intelligence, silent output generation increases the chance of unintended persistence, dissemination, or overwriting of artifacts.

External Transmission

Medium
Category
Data Exfiltration
Content
<div class="assignee-block">
  <div class="assignee-name">[target全称] <span class="tag">[角色：被诉/主张/共同申请人]</span></div>
  <div class="litigated-row">
    <div class="litigated-pn">🔴 <a href="https://analytics.zhihuiya.com/patent-view/abst'figures/?_type=query&source_type=search_result&rows=100&patentId=[真实patentId]">[CN号]</a></div>
    <div class="litigated-title">[专利真实标题]</div>
    <div class="family-chips">
      <span class="family-chip"><a href="[同族URL]">[同族公开号]</a> <span class="fc-auth">[受理局]</span> <span class="status-active">●有效</span></span>
Confidence
91% confidence
Finding
https://analytics.zhihuiya.com/

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.