ClawHub
Back to skill

Security audit

Research Idea

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed business-idea research helper that uses a background session, web search, local markdown output, and current-chat delivery in ways that match its purpose.

Install if you are comfortable with user-triggered background research sessions, web searches about your idea, persistent local files under ~/clawd/ideas/, and results being returned to the active chat. Use a private chat for confidential ideas and delete generated sessions or files when no longer needed.

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
92% confidence
Finding
The prompt explicitly instructs the agent to create a local directory and write a file under ~/clawd/ideas without surfacing that this skill performs filesystem modifications. In an agent setting, hidden write side effects are a real safety issue because a user asking only for idea research may not realize the skill will persist data locally, potentially exposing sensitive idea contents or causing unintended file creation.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.