voice-note-to-midi

Fail

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: HIGHCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: Modifies the user's ~/.bashrc file to append a custom directory to the system PATH, which serves as a persistence mechanism to ensure the skill's tools are always available.
  • [REMOTE_CODE_EXECUTION]: The installation process involves downloading and executing an external script (hum2midi) from an unspecified remote source, which may lead to the execution of unverified or malicious code.
  • [EXTERNAL_DOWNLOADS]: Fetches multiple third-party Python packages and dependencies, including basic-pitch and librosa, from public registries during setup.
  • [PROMPT_INJECTION]: Exposes a surface for indirect prompt injection through the ingestion and processing of untrusted audio files. Ingestion points: Processes WAV, M4A, and MP3 files for pitch extraction. Boundary markers: No markers or instructions are provided to the agent to ignore potentially malicious embedded content or metadata. Capability inventory: The skill performs file system writes and interacts with shell-level command execution. Sanitization: No input validation or sanitization steps are documented for the audio processing pipeline.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Mar 1, 2026, 05:16 AM