ClawHub
Back to skill

Security audit

Self Hosted Ai

Security checks across malware telemetry and agentic risk

Overview

This is a straightforward self-hosted AI deployment guide, but its server commands need careful review before use.

Install only if you want deployment guidance for servers you control. Before running the commands, review any downloaded install script, avoid exposing Ollama on 0.0.0.0 unless remote access is intentional, set strong N8N credentials, and use firewall rules, TLS, and a protected reverse proxy for internet-facing services.

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)

Vague Triggers

Medium
Confidence
93% confidence
Finding
The trigger phrases are broad deployment-related terms such as '装 N8N', '部署 Ollama', and '自己跑 AI', which can match ordinary user requests without clearly signaling high-impact system administration actions. This can cause the skill to activate in contexts where the user did not intend to authorize server changes, increasing the chance of unsafe or over-privileged guidance being applied.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill provides direct installation and system-modifying commands, including piping a remote script to shell, editing /etc/environment, restarting services, exposing Ollama on 0.0.0.0, and standing up internet-facing services, but it does not require confirmation, warn about authentication/network exposure, or instruct the user to validate prerequisites. In this self-hosting context, these steps can lead to unintended remote exposure, insecure defaults, or disruptive changes on production hosts if followed blindly.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal