ClawHub
Back to skill

Security audit

MeatLoop

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed human-review-by-email workflow, but users should avoid sending sensitive content to the external inbox.

Install only if you are comfortable with agents or users sending selected questions and attachments to an external Gmail inbox reviewed by a human. Do not submit credentials, private screenshots, regulated data, confidential documents, or legal, medical, or financial material.

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
96% confidence
Finding
The skill instructs users to send questions and attachments directly to a third-party Gmail address, which transfers potentially sensitive data outside the agent platform and its normal trust boundaries. Although the skill later mentions privacy limitations and advises not to submit confidential information, the external disclosure risk is not surfaced prominently before users are told where to send content, so agents or users may transmit screenshots, documents, or other sensitive material without adequate warning.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal