ClawHub
Back to skill

Security audit

Agent Skills Search

Security checks across malware telemetry and agentic risk

Overview

This is a simple Agnxi directory search skill that fetches a public sitemap and prints matching links, with minor transparency cautions about network use and broad auto-use wording.

Install this if you want your agent to search Agnxi for tools and skills. Be aware that using it makes an outbound request to agnxi.com, verify the package source because the README's clone URL is a placeholder, pass queries as safe process arguments rather than raw shell text, and review any returned tools before installing or using them.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
92% confidence
Finding
The skill declares no permissions, but its documented behavior invokes a local Python script that queries a live sitemap, which implies outbound network access. This creates a trust and transparency gap: agents or hosts may treat the skill as low-privilege while it can still reach external resources, potentially exposing prompts, queries, or enabling unexpected remote interaction.

Vague Triggers

Medium
Confidence
88% confidence
Finding
The README tells the agent to automatically use the `search_agnxi` tool whenever a user asks to find new tools, with broad examples like finding MCP servers or skills. This creates an overly general trigger that can overlap with ordinary user requests, causing the tool to be invoked without sufficient intent verification and potentially exposing the agent to unnecessary external lookups or prompt-scope expansion.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.