ClawHub
Back to skill

Security audit

visual-understanding

Security checks across malware telemetry and agentic risk

Overview

This skill is a straightforward GLM vision API guide that discloses sending selected images and prompts to Zhipu/BigModel, with no hidden execution or persistence found.

Install only if you intend to send images and prompts to Zhipu/BigModel services. Keep the API key in an environment variable, prefer an isolated Python environment, and avoid submitting confidential, regulated, or personally identifying images unless you have approval and understand the provider’s data handling terms.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

External Transmission

Medium
Category
Data Exfiltration
Content
---
name: glm-4.6v-vision-connector
description: "智谱 GLM-4.6V 多模态视觉模型集成插件。支持本地图像解析（Base64）及公网链接读取。优先提供 zai SDK 接入，并包含 cURL 原生降级方案。"
version: "1.1.0"
homepage: "https://github.com/zai-org/GLM-V"
repository: "https://github.com/zai-org/GLM-V.git"
Confidence
81% confidence
Finding
The skill is explicitly designed to send user-supplied images and prompts to an external third-party API endpoint, including public image URLs and Base64-encoded local images. This creates real data exfiltration risk if users submit sensitive images or confidential visual content, though the transmission is expected functionality rather than obviously malicious behavior.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.