ClawHub
Back to skill

Security audit

Beszel Check

Security checks across malware telemetry and agentic risk

Overview

This skill mostly does what it says, but it handles Beszel credentials in an under-protected way and includes an unexplained prompt to share server access with an unknown Gmail account.

Review before installing. Use a limited read-only Beszel account, avoid sourcing your full shell startup file for this skill, use only localhost or an HTTPS-protected trusted Beszel endpoint, and do not share your Beszel server with jenny@gmail.com unless you personally know and trust that account.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Missing User Warnings

Low
Confidence
93% confidence
Finding
The documented commands source `~/.zshrc` before running the skill, which executes arbitrary shell startup code from the user's local environment. If that file contains unsafe commands, has been tampered with, or performs unexpected side effects, invoking the skill can trigger unintended code execution unrelated to monitoring servers.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill accepts a host defaulting to an http URL and then sends the username and password to the authentication endpoint without enforcing TLS or warning the user. If BESZEL_HOST is set to a remote non-local endpoint or traffic leaves the host boundary, credentials can be exposed to interception or redirected to an attacker-controlled server.

VirusTotal

57/57 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.