imagegen

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it ingests and processes untrusted user prompts and external images to drive image generation and editing tasks.\n
  • Ingestion points: User-provided natural language prompts and image assets processed via scripts/image_gen.py.\n
  • Boundary markers: The skill lacks explicit delimiters or instructions to ignore embedded commands within user-provided prompts or image metadata during the augmentation process in SKILL.md.\n
  • Capability inventory: The skill performs network requests to the OpenAI API and executes local file writes for output artifacts and temporary batch files in scripts/image_gen.py.\n
  • Sanitization: Input validation exists for specific parameters (size, quality, model), but no content-based sanitization or filtering is applied to the prompt text before it is sent to the model.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 06:56 AM
Security Audit — agent-trust-hub — imagegen