sound-effects

Pass

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the user to install official packages from the vendor's managed registries, including the @elevenlabs/elevenlabs-js and @elevenlabs/cli packages from npm, and the elevenlabs package from PyPI. It also references a custom Scoop bucket and installation scripts hosted on the vendor's official GitHub repository.
  • [REMOTE_CODE_EXECUTION]: The installation guide includes a shell command pattern that downloads and executes an installer script (elevenlabs-cli-installer.sh) directly from the vendor's GitHub releases page using a pipe-to-shell (curl | sh) method.
  • [COMMAND_EXECUTION]: The skill provides examples for executing the elevenlabs CLI tool to perform sound conversions and authentication routines. It also demonstrates file system operations in Python and JavaScript to save generated audio data to local files.
  • [INDIRECT_PROMPT_INJECTION]: The skill identifies a surface for indirect prompt injection via the sound description input.
  • Ingestion points: The text parameter in the convert method accepts free-text descriptions of sound effects (SKILL.md).
  • Boundary markers: Input delimiters or specific "ignore embedded instructions" warnings are absent in the SDK usage examples.
  • Capability inventory: The skill performs network requests to the ElevenLabs API and writes binary data to the local disk (SKILL.md).
  • Sanitization: No explicit validation or sanitization of the input string is documented in the provided code examples.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 2, 2026, 04:37 AM
Security Audit — agent-trust-hub — sound-effects