ClawHub
Back to skill

Security audit

Wget Tool

Security checks across malware telemetry and agentic risk

Overview

This is a downloader skill with understandable purpose, but its documented command and features do not match the included code, and it can download untrusted content to local paths.

Install only if you are comfortable reviewing downloads and destinations yourself. Use trusted URLs, avoid passing long-lived secrets in headers, choose safe output directories, and do not assume the advertised wget-style options work unless the package adds a real, declared entrypoint and matching implementation.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Lp3

Medium
Category
MCP Least Privilege
Confidence
91% confidence
Finding
The skill clearly enables network access and likely shells out to a downloader, yet the metadata shown does not declare any permissions or safety boundaries. That gap is dangerous because users and policy systems may underestimate the skill’s ability to contact arbitrary remote hosts and fetch untrusted content.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill encourages use of custom headers and even shows an Authorization bearer token example without warning that these secrets will be transmitted to remote servers and may be logged, redirected, or exposed in histories. In a downloader context, this is especially risky because users may target arbitrary or compromised endpoints and unintentionally leak credentials.

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The script downloads arbitrary remote content and writes it directly to a user-specified path, or to a filename derived from the URL, without warning about destination choice or overwrite behavior. In an agent context, this can lead to unintended file replacement, writing into sensitive working directories, or planting attacker-controlled files that are later trusted or executed by a user or downstream workflow.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.