ClawHub
Back to skill

Security audit

AudioPod

Security checks across malware telemetry and agentic risk

Overview

AudioPod is a documentation-only skill for a third-party audio API, with expected API-key, billing, media-upload, and voice-cloning risks but no hidden or deceptive behavior found.

Install only if you are comfortable giving an agent an AudioPod API key that can create paid jobs, view wallet and usage metadata, and manage provider-side jobs or voices. Upload only media and voice samples you have permission to process, avoid sensitive recordings unless AudioPod's handling terms fit your needs, and verify the SDK package source before installing.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Description-Behavior Mismatch

Medium
Confidence
82% confidence
Finding
The skill documentation includes wallet and billing capabilities that are not reflected in the top-level manifest description, creating a scope mismatch. This can mislead users and reviewers about what the skill can access or do, especially since wallet endpoints expose account balance, pricing, and usage metadata beyond core audio processing.

Context-Inappropriate Capability

Low
Confidence
74% confidence
Finding
The wallet section grants access to balance, pricing, cost estimation, and usage history, which is not clearly necessary for the stated audio-processing purpose. While not directly code-executing, it expands the skill's data-access scope and may expose account or spending metadata that users did not expect to share.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The skill broadly advertises uploading local audio files and sending remote media URLs to a third-party API but does not warn about privacy, consent, retention, or third-party handling of potentially sensitive media. Audio, speech, and transcripts often contain personal or confidential information, so omission of such warnings increases the chance of unsafe use.

Missing User Warnings

High
Confidence
97% confidence
Finding
The documentation promotes voice cloning without any warning about consent, impersonation risk, or the handling of biometric voice data. Voiceprints are sensitive personal data, and enabling cloning without explicit safeguards materially increases the risk of fraud, deception, and privacy violations.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.