ClawHub
Back to skill

Security audit

wzz-server-monitor

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed server monitoring skill that installs a user-level cron job and sends configured SMTP alerts, with no evidence of hidden or destructive behavior.

Install only if you want ongoing local server monitoring. Before running install.sh, review the SMTP destination, keep the SMTP secret file permission-restricted, understand that alert emails include server identifiers and resource status, and inspect or remove the user crontab entry when you no longer need monitoring.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Rogue AgentSelf-Modification, Session Persistence
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (6)

Intent-Code Divergence

Medium
Confidence
94% confidence
Finding
The security section claims the script does not modify the system, but the same document instructs users to generate configuration files, write state/log files, and install cron entries. This misleading assurance can cause operators to underestimate persistence and system modification risks, especially when allowing an agent to execute setup steps automatically.

Context-Inappropriate Capability

Medium
Confidence
94% confidence
Finding
The installer automatically downloads and installs Python packages from an external PyPI mirror during setup. This expands the skill's behavior beyond simple server monitoring into network-based code acquisition and execution, creating supply-chain risk if the mirror, package resolution, or transport path is compromised. In this context, the behavior is somewhat expected for an installer, but it is still security-relevant because it executes newly fetched code on the host.

Missing User Warnings

Low
Confidence
91% confidence
Finding
The README describes SMTP-based alerting but does not prominently warn that monitoring data and alert content will be transmitted to an external mail server. In a monitoring skill, outbound notification is expected, but lack of explicit disclosure can still mislead users about data egress, especially when hostnames, resource state, and operational timing may be included in email alerts.

Vague Triggers

Medium
Confidence
86% confidence
Finding
The description says to use the skill whenever the user asks about server resource monitoring, load, or email alerts, which is broad trigger language without clear guardrails. In agent environments, this can cause over-activation and unexpected execution of monitoring or persistence-related actions in contexts where the user only wanted information, increasing the chance of unintended system changes.

Session Persistence

Medium
Category
Rogue Agent
Content
user-invocable: true
allowed-tools:
  - Read
  - Write
  - Bash(monitor.py *)
  - Bash(crontab *)
---
Confidence
93% confidence
Finding
Write - Bash(monitor.py *) - Bash(crontab *) --- # wzz-server-monitor — 服务器资源监控通知 监控 CPU / 内存 / 磁盘使用率，超过阈值时通过 SMTP 发送邮件告警。**核心脚本由 cron 独立定时运行，与任何 AI Agent 无关**；本 SKILL.md 是配置文档和交互入口。 ## 1. 快速开始

Session Persistence

Medium
Category
Rogue Agent
Content
echo "==> 5/5 安装 crontab（每 5 分钟检查一次）"
CRON_LINE="*/5 * * * * $PY $MONITOR check --config $CONFIG >> $STATE_DIR/monitor.log 2>&1"
( crontab -l 2>/dev/null | grep -v "$CRON_TAG" ; echo "# $CRON_TAG" ; echo "$CRON_LINE" ) | crontab -
echo "已写入 crontab:"
crontab -l | grep -A1 "$CRON_TAG" || true
Confidence
96% confidence
Finding
crontab -l

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.