ClawHub
Back to skill

Security audit

AI鞋履试穿效果图｜AI-HIVE

Security checks for vulnerabilities and agentic risk

Overview

This is a coherent AI-HIVE virtual shoe try-on helper with disclosed API use, media upload, billing cautions, and no evidence of hidden or destructive behavior.

Install only if you intend to use AI-HIVE for footwear preview generation. Provide only media you are authorized to upload, review parameters before any billable generation, and protect or remove the stored API key if you no longer use the tool.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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
Findings (4)

Lp3

Medium
Category
MCP Least Privilege
Confidence
90% confidence
Finding
The skill includes executable examples and operational guidance that use environment variables, local files, shell commands, and network access, but it declares no permissions or capability boundaries. This creates a transparency and governance gap: a host agent or reviewer may treat the skill as low-risk while it can still trigger external API calls, read/write local artifacts, and potentially incur charges or handle sensitive data.

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding
The documented purpose is narrowly framed as footwear try-on preview generation, but the described behavior and tooling appear capable of much broader actions such as generic text/video generation, arbitrary model enumeration, media upload, user/account queries, and local credential storage. This mismatch weakens user consent and policy enforcement because an orchestrator may invoke the skill assuming a narrow, low-risk function while it actually enables broader data access, billing, and content-generation behaviors.

Vague Triggers

Medium
Confidence
89% confidence
Finding
The default prompt invokes the skill through a broad, always-applicable phrasing ('Use $footwear-tryon-preview-ai-hive to turn my request...') without embedding narrow trigger conditions or scope checks. In combination with allow_implicit_invocation: true, this increases the chance the skill is auto-selected for loosely related requests and may cause unintended execution paths, policy bypass around user intent, or over-collection/processing of sensitive image inputs.

Natural-Language Policy Violations

High
Confidence
93% confidence
Finding
The default prompt hard-codes Chinese output ('production-ready Chinese workflow') regardless of the user's language preference. This can override user intent, reduce transparency for users who do not read Chinese, and create reviewability and safety-monitoring gaps if operators cannot easily inspect generated workflows, commands, or compliance-sensitive content.

Static analysis

No suspicious patterns detected.