Security audit

Elevenlabs

Security checks across malware telemetry and agentic risk

Overview

This is a coherent ElevenLabs audio skill, but users should treat API usage, text prompts, and voice-cloning samples as sensitive.

Install only if you are comfortable sending prompts, dialogue text, and selected audio samples to ElevenLabs using your API key. Keep the API key out of shared files or commits, and only clone voices when you have clear authorization from the speaker.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (5)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 95% confidence
Finding: The skill advertises and demonstrates capabilities that access environment variables, read/write files, invoke shell commands, and call an external network API, but it does not declare permissions accordingly. This creates a transparency and governance gap: users or orchestrators may grant or invoke the skill without understanding its true access scope, increasing the risk of unintended data exposure or unsafe execution.

Missing User Warnings

Medium

Confidence: 97% confidence
Finding: The description says the skill generates audio with ElevenLabs but does not clearly warn that user-provided text and possibly generated content are transmitted to a third-party API. Users may unknowingly send sensitive prompts or proprietary text off-platform, creating privacy, confidentiality, and compliance risk.

Missing User Warnings

High

Confidence: 99% confidence
Finding: The voice cloning section describes uploading voice samples to create clones but omits any warning that these samples are biometric identifiers and may be highly sensitive personal data. Without explicit consent and privacy warnings, users may upload third-party voices or their own recordings without understanding impersonation, consent, retention, and regulatory risks.

Missing User Warnings

Medium

Confidence: 81% confidence
Finding: The function transmits user-provided dialogue text, voice selections, and optional language/seed metadata to a third-party API without any in-band consent prompt or data sensitivity check. In an agent skill context, this can cause unintended disclosure of sensitive user content if callers are unaware that prompts are sent off-platform.

Missing User Warnings

Medium

Confidence: 81% confidence
Finding: This function uploads user-provided audio samples to ElevenLabs, which may contain biometric voice data and sensitive personal information, but the code provides no explicit in-tool warning or consent checkpoint before transmission. In an agent/skill context, this is more dangerous because users may invoke the skill indirectly and may not realize local files are being sent to a third-party service.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.