hatch-pet

Pass

Audited by Gen Agent Trust Hub on Jun 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill coordinates several local Python scripts for deterministic image processing using the subprocess module. In finalize_pet_run.py and render_animation_videos.py, it executes bundled tools to extract frames, compose the atlas, and render preview videos using ffmpeg. All commands are restricted to local scripts or standard system utilities.
  • [EXTERNAL_DOWNLOADS]: The generate_pet_images.py script performs network operations to communicate with the OpenAI API (api.openai.com) for its secondary image generation fallback. This activity is restricted to a well-known service and is necessary for the skill's primary function of creating character art.
  • [PROMPT_INJECTION]: The skill processes untrusted user data, such as pet descriptions and reference images, which are interpolated into prompts for the $imagegen system skill. This constitutes an indirect prompt injection surface. However, the skill provides mitigation through input sanitization (e.g., slugification in prepare_pet_run.py) and manual verification steps before finalization.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 28, 2026, 07:23 PM