ClawHub
Back to skill

Security audit

Llm Data Automation

Security checks across malware telemetry and agentic risk

Overview

This skill appears to be a coherent local data-automation helper, with filesystem access and document indexing aligned to its stated purpose.

Before installing, treat it like a local data-processing assistant with access to files you point it at. Use approved local/private models for confidential documents, avoid indexing broad company folders unless you have permission, and review generated scripts before running them on important data.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Confidence
91% confidence
Finding
This section instructs users to load all PDFs from a company_documents folder into a queryable index, which could include sensitive internal documents. Although the file later suggests using local LLMs for sensitive data, this specific workflow lacks a direct warning to review document sensitivity, access controls, or data handling implications before indexing and querying company content.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.