elevenlabs-music

Pass

Audited by Gen Agent Trust Hub on Jul 22, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: References and suggests installation of the belt-sh/cli package and additional skills from the inference-sh/skills repository on GitHub. These are the official tools for the service being integrated.
  • [COMMAND_EXECUTION]: Utilizes the belt CLI tool to perform music generation. The execution environment is restricted by the allowed-tools frontmatter to only permit the belt command, providing a strong layer of defense against arbitrary command execution.
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it interpolates user-provided text directly into a CLI command. Ingestion points: The prompt and duration_seconds parameters used in the SKILL.md examples. Boundary markers: The JSON payload is enclosed in single quotes within the bash command. Capability inventory: Access is restricted to the Bash tool, specifically limited to the belt binary as defined in the frontmatter. Sanitization: No explicit sanitization or escaping of the user-provided prompt is performed within the instruction file.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 22, 2026, 07:32 AM
Security Audit — agent-trust-hub — elevenlabs-music