ClawHub

Agent Skills Search

Security checks across malware telemetry and agentic risk

Overview

This skill is a simple Agnxi directory search helper that fetches a public sitemap and prints matching links, with minor disclosure and provenance cautions.

Reasonable to install if you want an agent to search Agnxi. Be aware it will contact Agnxi.com when used, verify the publisher/source because the README has a placeholder repository URL, pass queries as safe process arguments rather than raw shell text, and review any returned tools or MCP servers 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
90% confidence
Finding
The skill invokes a local Python script that performs live queries against an external sitemap, which implies network access, but the skill declares no corresponding permissions or trust boundary. This creates a transparency and policy-enforcement gap: agents or hosts may execute network-capable behavior users did not explicitly approve, increasing the risk of unexpected outbound requests, data exposure through query contents, or misuse in restricted environments.

Vague Triggers

Medium
Confidence
88% confidence
Finding
The README states that the agent will utilize the `search_agnxi` tool automatically when asked to find new tools, and the example prompts are broad enough to overlap with normal user requests. This can cause over-triggering, where the skill activates on common discovery queries without clear user consent or stronger scoping, increasing the chance of unnecessary external requests and unintended workflow influence.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal