storyteller-image-maker

Pass

Audited by Gen Agent Trust Hub on Feb 27, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill uses phrases such as 'IMPORTANT' and 'CRITICAL' within its generated prompts. However, these are identified as prompt engineering techniques directed at the image generation model (Gemini) to ensure visual quality and prevent the generation of text within images. They do not attempt to override the AI agent's own safety guidelines or system instructions.
  • [INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting data from an external source.
  • Ingestion points: The skill reads image-concept and Template data from the planner agent's output.
  • Boundary markers: The skill uses a structured prompt template to wrap the external content between style prefixes and composition suffixes, which helps isolate the external data.
  • Capability inventory: The skill utilizes the nanobanana__generate_image tool to produce image files.
  • Sanitization: No explicit sanitization or filtering of the image-concept text is performed before it is interpolated into the final prompt sent to the image generation tool.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 27, 2026, 01:00 AM