ClawHub
Back to skill

Security audit

image-prompting

Security checks across malware telemetry and agentic risk

Overview

This is a coherent image-prompting skill with visible external API examples, but users should treat persona and avatar examples as sending prompts, image URLs, and scripts to third-party services.

Installers should understand this is mostly prompt documentation. If using the curl examples or related Pruna/Replicate workflows, do not include secrets, confidential assets, internal URLs, or real-person images/voices without permission, and review the provider's handling of uploaded media and scripts before use.

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 (4)

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The example includes an authenticated curl request to an external API using prompts and model inputs, but it does not warn users that their prompt content and related metadata will be transmitted to a third-party service. In a skill focused on persona generation, users may paste sensitive descriptions, internal image URLs, or proprietary creative material into these examples without realizing the data leaves their environment.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The avatar examples upload image references, voice scripts, voice selections, and video prompt data to an external service, but the documentation does not disclose that this content is transmitted off-platform. Because this skill is specifically about realistic personas and avatar creation, the omitted warning is more sensitive: users may submit personal likenesses, identity-bearing images, or sensitive scripts assuming the examples are local-only.

External Transmission

Medium
Category
Data Exfiltration
Content
-H 'Model: p-video-avatar' \
  -d '{
    "input": {
      "image": "https://api.pruna.ai/v1/files/APPROVED_STILL_ID",
      "voice_script": "Hey — we put the patchwork set on a real street plate, not a white studio. Face and background stayed put.",
      "voice": "Puck (Male)",
      "voice_language": "English (US)",
Confidence
91% confidence
Finding
https://api.pruna.ai/

External Transmission

Medium
Category
Data Exfiltration
Content
-H 'Model: p-video-avatar' \
  -d '{
    "input": {
      "image": "https://api.pruna.ai/v1/files/ANIME_STILL_ID",
      "voice_script": "So — same motion grammar, totally different world. That is the point of planning each scene.",
      "voice": "Zephyr (Female)",
      "voice_language": "English (US)",
Confidence
91% confidence
Finding
https://api.pruna.ai/

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.