minimax-h3-video-toolkit-ai-hive

Pass

Audited by Gen Agent Trust Hub on Sep 23, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The script scripts/edit_video.py uses subprocess.run to call ffmpeg and ffprobe for local media editing tasks like aspect ratio changes and duration trimming. This is a legitimate functional requirement and is implemented using safe argument handling.
  • [EXTERNAL_DOWNLOADS]: scripts/videogen.py performs network requests to ai-hive.iclip.cn for media uploads and generation tasks. This activity is restricted to the vendor's API and is necessary for the skill's core functionality.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied creative prompts and media assets as part of the generation workflow. While this constitutes a data ingestion surface for potential indirect prompt injection, it is a standard characteristic of generative video tools and is managed by the service's API safety layers.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 23, 2026, 02:49 PM