ClawHub
Back to skill

Security audit

Stable Diffusion 图片生成替代｜AI 图片生成与编辑

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed AI Hive image-generation helper that uses a fixed remote image model and optional user-provided reference images.

Install only if you intend to use AI Hive's remote image API. Running generation commands can upload the reference images you provide, use an API key, submit potentially billable jobs, and save outputs under ~/Downloads/AiHive. Keep the stored API key private and remove ~/.ai-hive/config.json if you no longer want local credential persistence.

SkillSpector

Vulnerability Patterns
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
86% confidence
Finding
The skill advertises executable commands and operational capabilities including shell execution, filesystem access, environment-variable use, and outbound network access, but it does not declare permissions. This creates a transparency and trust problem: users and enforcement layers cannot accurately assess what the skill can do, and hidden capability scope increases the risk of unauthorized key handling, file modification, or network exfiltration if the underlying script is unsafe.

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding
The declared purpose says the skill only helps migrate Stable Diffusion-style workflows to AI Hive constraints, but the behavior includes API key initialization and local storage, direct image-generation API calls, model/pricing queries, media upload, task polling, and result download. This mismatch is dangerous because users may consent to a seemingly narrow transformation tool while the skill actually performs credential handling, remote uploads, and potentially billable network actions with persistent local state.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.