Text to Speech and Voice Cloning by Verbatik AI

• Outcome: Review
• Latest audit: May 29, 2026, 9:23 PM
• Version:
  1.0.1

Overview

The skill is a coherent Verbatik text-to-speech integration, but it gives an agent sensitive voice-cloning and deletion authority without built-in consent or confirmation safeguards.

Review before installing. Use this only if you trust Verbatik and the publisher, keep the API key scoped and protected, require explicit confirmation before spending balance, storing or sharing generated audio, cloning a voice, or deleting a cloned voice, and only clone voices when you have documented permission from the speaker.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (2)

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The skill enables voice cloning from arbitrary audio without any warning or guardrails around consent, impersonation, biometric privacy, or authorization to use the source voice. In a voice-cloning context, omitting these safeguards materially increases the risk of fraud, impersonation, and misuse of sensitive biometric data.

Missing User Warnings

Low

• Confidence: 76% confidence
• Finding: The documented delete operation is destructive, but the skill provides no caution that deleting a cloned voice may be irreversible or operationally disruptive. This can lead to accidental loss of assets or interruption of downstream workflows that depend on the voice ID.

VirusTotal

62/62 vendors flagged this skill as clean.