ClawHub
Back to skill

Security audit

Bring Your Own Model

Security checks across malware telemetry and agentic risk

Overview

This skill is a straightforward Runware model-import guide, but users should protect any proprietary model file they upload.

Before installing or using this skill, treat model weights as sensitive intellectual property: use a signed, time-limited URL rather than a broadly public file, keep the Runware model private unless you intend to publish it, and record the AIR and uniqueIdentifier so you can update or delete the model later.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The skill instructs users to provide a publicly reachable or signed download URL for model weights and emphasizes server-side fetching, but it does not warn about the privacy, retention, or unintended disclosure risks of exposing proprietary model files through such URLs. In this context, uploaded weights may contain sensitive intellectual property or licensed content, so encouraging public accessibility without safeguards can lead to accidental exposure or unauthorized redistribution.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.