ClawHub
Back to skill

Security audit

A股逻辑可视化工具

Security checks across malware telemetry and agentic risk

Overview

The skill is a coherent finance-diagram helper, but it writes HTML files to arbitrary paths and those files load a third-party Draw.io viewer script when opened.

Review this before installing if you want generated diagrams to stay fully local. Prefer XML output or constrain HTML output to a workspace directory unless you are comfortable opening files that load the diagrams.net viewer from the internet.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (3)

Lp3

Medium
Category
MCP Least Privilege
Confidence
85% confidence
Finding
The skill explicitly instructs use of `VisualizerTools.render_drawio_to_html(xml_content, filename)`, which implies writing an HTML artifact to disk, yet the skill declares no permissions. Undeclared file-write capability weakens security review and runtime enforcement because consumers may invoke a skill that creates files without that capability being transparently declared.

Context-Inappropriate Capability

Medium
Confidence
89% confidence
Finding
The generated HTML loads a remote script from viewer.diagrams.net, creating a client-side supply-chain dependency at render time. If that third-party script is compromised, blocked, or replaced, anyone opening the produced HTML could execute attacker-controlled JavaScript in their browser, which is especially risky because the HTML also embeds user-provided diagram XML.

Vague Triggers

Medium
Confidence
81% confidence
Finding
The trigger language is broad: it activates when users ask to 'visualize, draw, map, diagram, or explain complex finance logic visually.' Overbroad routing can cause the skill to handle requests outside its narrowly intended scope, increasing the chance that users are routed into file-generating or code-adjacent behavior unnecessarily and bypass more appropriate, safer skills.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.