ClawHub
Back to skill

Security audit

glm-grounding

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only image grounding skill that uses a model endpoint as described and does not include hidden code or installation behavior.

Before installing, confirm that the configured GLM endpoint is trusted and that any images you provide are appropriate to send there. Avoid using sensitive images unless you understand the endpoint's logging, retention, and access controls.

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)

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill documentation instructs sending user-supplied images to a model API endpoint, including an internal HTTP service, without any explicit privacy notice, consent step, or data-handling disclosure. If users provide sensitive images, this can result in unintended transmission of personal or confidential data to backend services, especially since the examples normalize use of remote inference without surfacing transfer risks.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.