claude-video-enhance-audio

Pass

Audited by Gen Agent Trust Hub on Sep 20, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted media files to generate transcripts and speaker diarization data. This creates an attack surface where an attacker could embed spoken instructions in an audio file that, when transcribed, might influence the agent's behavior.
  • Ingestion points: The skill uses scripts/audio_enhance.py to ingest external video and audio files ($INPUT) for processing with WhisperX and other AI models.
  • Boundary markers: The instructions do not specify any boundary markers or delimiters for the generated transcripts, nor do they instruct the agent to ignore any commands found within the processed content.
  • Capability inventory: The agent has access to powerful tools including Bash for command execution and Read/Write for filesystem access, which could be abused if an injection is successful.
  • Sanitization: There is no evidence of sanitization or filtering of the output text (such as speakers.json or transcript segments) before it is passed back to the agent's context.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 20, 2026, 06:38 PM
Security Audit — agent-trust-hub — claude-video-enhance-audio