Skills
Navigate
skills/styleof/superpowers/nano-banana-2/Gen Agent Trust Hub

nano-banana-2

Pass

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow where user-controlled strings (prompts) and external image URLs are passed directly to an AI model and CLI tool.
  • Ingestion points: The prompt and images parameters in the --input JSON payload for the belt CLI command in SKILL.md.
  • Boundary markers: None; the user input is interpolated directly into JSON structures without delimiters or instruction isolation.
  • Capability inventory: The skill is configured to use the Bash tool specifically restricted to the belt command (allowed-tools: Bash(belt *)).
  • Sanitization: The skill instructions do not include methods for validating or sanitizing the content of the prompts or the sourced image URLs.
  • [EXTERNAL_DOWNLOADS]: The skill references and provides instructions for installing external software associated with the inference.sh platform.
  • Evidence: Instructions to install the CLI via npx skills add belt-sh/cli and references to a Python SDK (inferencesh).
  • Evidence: External link to installation guidelines hosted on GitHub (https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md).
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 2, 2026, 02:21 PM
Security Audit — agent-trust-hub — nano-banana-2