skills/zju-real/easel/audio-mix/Gen Agent Trust Hub

audio-mix

Pass

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructions direct the agent to execute a local script located at skills/shared/scripts/audio_mix.py using the Python interpreter. This script is used to perform audio mixing tasks such as ducking and SFX overlay.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted audio data from external files, which could potentially contain malicious instructions or malformed metadata intended to exploit the processing logic. 1. Ingestion points: Audio files provided as inputs for voice, background music, and sound effects in SKILL.md. 2. Boundary markers: None present in the skill instructions to separate data from commands. 3. Capability inventory: Subprocess execution via python for audio processing tasks as seen in the execution steps. 4. Sanitization: No explicit sanitization or validation of the input audio files is described in the provided instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 2, 2026, 01:21 PM