ClawHub
Back to skill

Security audit

Agnes Image Generator

Security checks across malware telemetry and agentic risk

Overview

This skill is a straightforward Agnes image-generation wrapper, but users should know prompts, reference images, and generated outputs are sent to an external service and returned as public cloud URLs.

Install only if you are comfortable sending prompts and reference images to Agnes and receiving public cloud-hosted image URLs. Use an environment variable for the API key, avoid hardcoding credentials, and do not submit confidential, personal, regulated, or private images unless you accept the external processing and public-output behavior.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The README states that generated images are returned as cloud-hosted URLs and notes they are publicly accessible, but it does not clearly warn users about the privacy and confidentiality implications of uploading prompts or source images that may contain sensitive data. In an image-generation skill, this can lead users to unknowingly expose private images or sensitive derived content to anyone with the URL or via unintended sharing/logging paths.

Vague Triggers

Medium
Confidence
94% confidence
Finding
The trigger condition is broad enough to activate on essentially any image-generation-related request, which can cause the skill to run outside a narrowly scoped invocation boundary. In practice, this increases the chance that user prompts and optional image inputs are sent to the external Agnes API unintentionally, creating privacy, consent, and tool-routing risks.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The skill describes accepting prompts and reference image URLs/Data URIs and sending them to a third-party API, but it does not clearly warn users that this data leaves the local system. This omission can lead to unintentional disclosure of sensitive prompts, private image URLs, or embedded image data to an external service without informed consent.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal