ClawHub
Back to skill

Security audit

GPT-SoVITS TTS

Security checks across malware telemetry and agentic risk

Overview

This TTS skill does what it claims, but it has a real command-injection risk in its MP3 conversion path and needs review before use.

Review before installing. Only use reference audio you have permission to clone, avoid sensitive text unless you trust the configured GPT-SoVITS endpoint, and do not pass user-controlled output paths until the ffmpeg call is changed to a safer argument-array API such as execFile or spawn with path validation.

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
78% confidence
Finding
The skill sends arbitrary input text to an external/local TTS API over HTTP without any disclosure, minimization, or trust validation. If users provide sensitive data, that content is transmitted off-module and could be logged, intercepted, or processed by an unintended service, especially because the endpoint is configurable via environment variable.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

Detected: suspicious.dangerous_exec

Shell command execution detected (child_process).

Critical
Code
suspicious.dangerous_exec
Location
index.js:57