create-cartoon-music-video

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill extensively uses ffmpeg for complex media processing tasks, including video concatenation, logo bug overlays, and audio-video muxing. It also executes local Python scripts (stitch.py, rebuild_captions.py) and shell scripts to automate the production pipeline. These are standard operations for media orchestration.
  • [EXTERNAL_DOWNLOADS]: The skill interacts with established AI platforms (ElevenLabs, Higgsfield, Google) to generate assets. It also incorporates a final delivery step that uploads video content to a vendor-managed ads library API via authenticated requests. These operations are core to the skill's functionality and target appropriate infrastructure.
  • [PROMPT_INJECTION]: Since the skill ingests user-supplied lyrics and aesthetic descriptions to generate AI content, it possesses an inherent surface for indirect prompt injection. However, it implements basic sanitization by stripping section markers from lyrics and follows a structured workflow that mitigates the risk of unexpected behavior during generation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 03:07 PM
Security Audit — agent-trust-hub — create-cartoon-music-video