ClawHub
Back to skill

Security audit

Poetry Recitation

Security checks across malware telemetry and agentic risk

Overview

This skill appears to do what it says: it creates poem-recitation videos using a local TTS pipeline and available voices, with no hidden persistence or unrelated behavior found.

Install only if you trust the local TTS pipeline it depends on and you have permission to use any cloned voices available there. Review the poem text, selected voice, and output path before running when voice identity or consent matters.

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 (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
95% confidence
Finding
The skill invokes shell commands and reads local files or directories, but declares no permissions or trust boundaries. That creates a transparency and policy gap: an agent may execute filesystem and shell actions users and reviewers did not explicitly authorize, increasing the risk of unintended local access or command misuse.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The workflow instructs automatic delivery of generated media with 'no extra confirmation needed' even when cloned voices may be used. Because cloned voice output can impersonate real people or produce sensitive likeness-based content, skipping a confirmation or consent checkpoint raises the chance of unauthorized impersonation, privacy violations, or accidental distribution of harmful media.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.