ClawHub
Back to skill

Security audit

Mxe

Security checks across malware telemetry and agentic risk

Overview

This Markdown export skill is useful, but its setup asks the agent to build and globally link unreviewed local or external Node tooling.

Install only if you can inspect and trust the actual `mxe` project at the referenced local path, or replace the setup with a reviewed, pinned package. Expect URL inputs to make outbound web requests, and remember that local images referenced by Markdown may be embedded into exported documents.

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 (2)

Missing User Warnings

Low
Confidence
86% confidence
Finding
The skill explicitly encourages fetching arbitrary remote URLs and converting them without warning about outbound network access, privacy implications, or trust boundaries. In an agent setting, this can cause unintended requests to third-party or internal resources, exposing IP/network metadata or enabling retrieval of sensitive intranet content if the tool is used on attacker-supplied URLs.

Vague Triggers

Medium
Confidence
91% confidence
Finding
The trigger list includes generic phrases such as "export pdf", "export document", and "generate pdf", which can cause the skill to activate for user requests that are not specifically about this markdown conversion tool. Overly broad invocation phrases increase the chance of unintended execution, confusing tool selection, and accidental processing of content in contexts where a more appropriate skill should be used.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.