ClawHub
Back to skill

Security audit

ace-music

Security checks for vulnerabilities and agentic risk

Overview

This skill is a noisy but mostly straightforward music-generation integration that uses an external ACE Music API and may create local audio files.

Install only if you are comfortable sending prompts, lyrics, and generation parameters to ACE Music. Keep the API key in a private environment variable instead of shared project files, review any generated commands before running them, and avoid enabling broad request-content logging.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Vague Triggers

High
Confidence
90% confidence
Finding
The trigger keywords are extremely broad and noisy, including generic terms such as 'full', 'open-source', 'step', and malformed fragments. In an agent environment, this can cause unintended invocation of the skill on unrelated prompts, leading to unexpected external API usage, command execution, or file generation without clear user intent.

Missing User Warnings

Medium
Confidence
87% confidence
Finding
The skill advertises execution of commands and file-handling behavior but does not provide a prominent user-facing warning or consent boundary for system-impacting actions. In an agent context, that increases the risk of surprising local command execution, file creation, or interaction with external services triggered by ordinary content requests.

Ssd 3

Medium
Confidence
97% confidence
Finding
The guidance recommends logging full API request content, which may contain user prompts, lyrics, uploaded-derived content references, or other sensitive material. If such logs are stored broadly or retained long term, they can leak private user data, copyrighted material descriptions, or tokens embedded in request metadata.

Static analysis

No suspicious patterns detected.