gpt-image
Pass
Audited by Gen Agent Trust Hub on Nov 13, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references and provides instructions to install external tools and additional skill sets from the belt-sh and inference-sh repositories. These are the official sources for the platform used by the skill.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes external data via prompt, images, and mask parameters which are passed as arguments to the belt CLI tool. This represents a potential surface for indirect prompt injection if malicious content is retrieved from external sources.\n
- Ingestion points: User-provided strings and URLs in the prompt, images, and mask fields within SKILL.md.\n
- Boundary markers: The instructions do not define specific delimiters or ignore instructions for the interpolated data.\n
- Capability inventory: The skill uses the Bash(belt *) tool to execute commands.\n
- Sanitization: There is no explicit sanitization described in the skill for the input parameters before they are passed to the shell.
Audit Metadata