voice-isolator

Pass

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The installation guide instructs users to download official CLI tools and libraries from the vendor's verified infrastructure, including NPM, PyPI, Homebrew, and Scoop repositories.- [REMOTE_CODE_EXECUTION]: The documentation includes a shell installation method that pipes a remote script from the vendor's official GitHub repository directly into the shell.- [DATA_EXFILTRATION]: Code examples in the documentation demonstrate fetching audio data from external URLs using the requests library, which is a standard part of the skill's functionality for processing remote audio files.- [INDIRECT_PROMPT_INJECTION]: The skill demonstrates processing content from external URLs without explicit validation in the illustrative code examples, which presents a technical ingestion surface for untrusted data.
  • Ingestion points: External audio URLs fetched via requests.get in SKILL.md.
  • Boundary markers: Not included in the provided code snippets.
  • Capability inventory: Network requests, local file system writing, and specialized API interaction.
  • Sanitization: Not addressed in the basic illustrative examples.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 2, 2026, 04:38 AM
Security Audit — agent-trust-hub — voice-isolator