nano-banana-2
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill requires the installation of the 'belt' CLI tool and suggests adding related skills from the 'inference-sh' GitHub organization. It also references a Python SDK named 'inferencesh'.
- [COMMAND_EXECUTION]: The skill uses the 'Bash' tool to execute CLI commands such as 'belt login', 'belt app run', and 'belt app sample' to interact with the inference.sh platform.
- [INDIRECT_PROMPT_INJECTION]: The skill instructions show the agent how to build shell commands by interpolating user-provided prompt text into a JSON string argument for the CLI.
- Ingestion points: The 'prompt' parameter in the input JSON object found in SKILL.md.
- Boundary markers: User input is wrapped in JSON keys and enclosed in single quotes for the shell command.
- Capability inventory: The skill is granted access to the 'Bash' tool with 'belt' commands.
- Sanitization: No explicit sanitization or validation logic is provided in the instructions to prevent command injection via malicious input characters.
Audit Metadata