ClawHub
Back to skill

Security audit

agent4.io

Security checks across malware telemetry and agentic risk

Overview

The skill appears purpose-aligned, but its installer asks users to run a live remote shell script that changes local agent configuration.

Review the installer before running it, prefer manual MCP setup when possible, and use an agent4.io API key with only the access you are comfortable giving to local agent tooling. Be careful when importing private documents or publishing agents and Storylines because those changes persist in the tenant.

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
90% confidence
Finding
The skill instructs users to run a remote shell script via `curl ... | sh`, which executes code fetched over the network without inspection. Even if the vendor is legitimate, this pattern is dangerous because it can modify local agent/client configuration and becomes a code-execution vector if the endpoint, distribution path, or account is compromised.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.