Skills
skills/inference-shell/skills/nano-banana-2/Gen Agent Trust Hub

nano-banana-2

Pass

Audited by Gen Agent Trust Hub on Apr 22, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references installation instructions for the CLI tool from the vendor's official GitHub repository at github.com/inference-sh/skills.
  • [COMMAND_EXECUTION]: The skill utilizes the infsh CLI tool, which is the designated interface for the inference-shell platform. Its usage is consistent with the skill's stated purpose of running image generation models.
  • [PROMPT_INJECTION]: The skill incorporates user-provided text prompts and image URLs into the model's input payload. This represents a standard surface for indirect prompt injection, typical for generative AI tools.
  • Ingestion points: prompt and images fields in command arguments and SDK calls within SKILL.md.
  • Boundary markers: Not present; user input is interpolated directly into JSON structures.
  • Capability inventory: The skill is scoped to the Bash(infsh *) tool for model execution.
  • Sanitization: No explicit filtering or sanitization of input text is described in the instruction set.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 22, 2026, 04:25 PM
Security Audit — agent-trust-hub — nano-banana-2