p-image
Pass
Audited by Gen Agent Trust Hub on Jul 7, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides integration for the inference.sh platform using its official belt CLI tool. All referenced domains and tools are consistent with the skill's stated purpose of optimized image generation.
- [EXTERNAL_DOWNLOADS]: The skill documentation includes a link to installation instructions hosted on the official inference-sh GitHub repository.
- Evidence: Reference to https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md for CLI setup instructions.
- [PROMPT_INJECTION]: The skill processes user-provided text prompts, creating a surface for potential indirect prompt injection attacks.
- Ingestion points: User prompts are passed via the --input flag to the belt CLI tool in SKILL.md.
- Boundary markers: Prompts are encapsulated within JSON objects passed as command arguments.
- Capability inventory: The skill is restricted to executing the belt CLI tool via the Bash(belt *) tool constraint.
- Sanitization: Input data is structured within a JSON payload, which provides separation from the underlying shell command logic.
Audit Metadata