ClawHub

Audio Editor

Security checks across malware telemetry and agentic risk

Overview

This audio-editing skill mostly matches its stated purpose, but its shell helper can execute arbitrary commands instead of staying limited to audio processing.

Review before installing. Do not run this version as-is unless it is changed to remove eval, reject unsupported natural-language requests, validate paths and numeric values, and invoke ffmpeg through safely quoted arguments.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Context-Inappropriate Capability

High
Confidence
99% confidence
Finding
The script treats unrecognized natural-language input as a shell command and then executes it via eval. This allows arbitrary command execution completely outside the audio-processing scope, so any caller controlling COMMAND can run destructive or data-exfiltrating shell operations.

Missing User Warnings

High
Confidence
98% confidence
Finding
The script executes a parsed command string with eval without any confirmation, escaping, or constrained command construction. Even within intended flows, this magnifies injection risk because filenames and parsed values are interpolated into a shell command string that the shell re-parses.

VirusTotal

48/48 vendors flagged this skill as clean.

View on VirusTotal