hatch-pet

Pass

Audited by Gen Agent Trust Hub on Nov 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: User-provided character descriptions are interpolated into prompts for external image generation services. This creates an indirect prompt injection surface; however, the risk is inherent to the skill's primary function of text-to-image generation and is limited to the visual content of the produced assets.
  • [EXTERNAL_DOWNLOADS]: The skill communicates with image generation APIs (such as OpenAI or Kling AI) to generate and download sprite strips. These network operations are a core part of the skill's functionality and are performed via user-configured API endpoints.
  • [COMMAND_EXECUTION]: The generate.py script executes the skill's own prepare.py script using subprocess.run to handle specific chroma-key conflicts and recalculate prompt assets. This is a controlled execution of local scripts included within the skill package.
  • [CREDENTIALS_SAFE]: The skill facilitates the storage of API keys in a local .env file and provides explicit instructions for the agent to mask these keys in logs and exclude them from version control, adhering to standard security practices for local development.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 7, 2026, 08:01 AM