ClawHub
Back to skill

Security audit

AI大模型专家｜企业AI网关与成本路由

Security checks across malware telemetry and agentic risk

Overview

The skill is a disclosed AI-HIVE gateway/media workflow helper that stores an API key locally and runs user-invoked API calls, with no evidence of hidden exfiltration or destructive behavior.

Install only if you intend to use AI-HIVE with your own or authorized API key. Treat generated commands as actions that may upload selected local media, submit billable model tasks, query task status, and save outputs locally. Review when implicit invocation is acceptable for your workspace.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (3)

Lp3

Medium
Category
MCP Least Privilege
Confidence
86% confidence
Finding
The skill advertises and invokes capabilities including environment access, file read/write, network calls, and shell execution, but does not declare any permissions. This creates a trust and review gap: users and platforms cannot accurately assess what the skill may access before use, increasing the risk of unintended credential exposure, local file interaction, or command execution through bundled scripts.

Tp4

High
Category
MCP Tool Poisoning
Confidence
89% confidence
Finding
The documented purpose is an enterprise AI gateway and routing assistant, but the referenced behavior extends into direct model invocation, media generation, credential setup, local credential storage, balance retrieval, and unrelated blueprint generation. This mismatch is dangerous because users may trust the skill for architecture guidance while it actually performs broader sensitive actions, including handling API keys and making live account-affecting requests.

Vague Triggers

Medium
Confidence
93% confidence
Finding
The manifest sets allow_implicit_invocation to true without any visible trigger constraints, which can cause the skill to activate in broader contexts than the user explicitly intended. For a skill that influences enterprise AI gateway design, routing, auditing, and executable command generation, accidental invocation could lead to unintended policy guidance, operational changes, or sensitive workflow shaping.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.exposed_secret_literal

File appears to expose a hardcoded API secret or token.

Critical
Code
suspicious.exposed_secret_literal
Location
SKILL.md:126