ClawHub
Back to skill

Security audit

Dlazy Jimeng Omnihuman 1.5

Security checks across malware telemetry and agentic risk

Overview

The skill is purpose-aligned, but it needs review because the pinned CLI can upload selected local media even during dry-run and stores API keys without the permission hardening the skill claims.

Review before installing. Prefer npx if you do not need a persistent global CLI, avoid using dry-run with sensitive local images or audio unless this behavior is fixed or confirmed acceptable, and consider using DLAZY_API_KEY per invocation or manually locking down ~/.dlazy/config.json permissions if you save an API key.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Vague Triggers

Medium
Confidence
82% confidence
Finding
The trigger keywords are broad enough to overlap with ordinary user requests, which can cause the skill to activate unexpectedly. In this skill, unintended activation is more sensitive because it is wired to an external CLI that can upload local image/audio files to remote services and requires credentialed API access, increasing the chance of accidental data egress or unintended paid API usage.

Vague Triggers

Medium
Confidence
85% confidence
Finding
The trigger keyword "digital human" is very broad and can match many ordinary user requests unrelated to this specific vendor tool, causing the skill to activate unexpectedly. Over-broad activation can route benign conversations into a third-party CLI/API workflow, increasing the chance of unintended data disclosure, unwanted tool execution, or user confusion about what is being invoked.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.