ClawHub
Back to skill

Security audit

GPT-Image-2 for OpenClaw

Security checks across malware telemetry and agentic risk

Overview

This skill coherently installs an OpenClaw image-generation provider, with the main caution that reinstalling overwrites the existing local hnbc provider folder.

Install this only if you intend to add or replace the hnbc image-generation provider. Back up any custom files under ~/.openclaw/extensions/hnbc first, keep the HNBC API key private, and use a trusted HNBC base URL because prompts and credentials are sent to that service during generation.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
93% confidence
Finding
The skill directs the agent to run shell scripts, inspect runtime state, and configure a network-backed image provider, which implies environment and network access, but the skill metadata does not declare those capabilities. This mismatch can cause users or orchestrators to approve and execute a skill without understanding that it will read configuration/auth context and interact with external services, weakening least-privilege and review controls.

Missing User Warnings

Medium
Confidence
87% confidence
Finding
The script unconditionally removes the destination plugin directory with 'rm -rf' before copying new contents, without warning, confirmation, backup, or validation that the target contains only the expected plugin files. In an install context this is often intended behavior, but it is still destructive and can cause data loss if the destination path is misconfigured or if users have local modifications stored under that directory.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.