p-image
Pass
Audited by Gen Agent Trust Hub on Nov 13, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Suggests installation of the belt command-line tool and references external installation scripts and documentation hosted on GitHub and the vendor's domain. * Evidence: npx skills add belt-sh/cli and https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md.
- [COMMAND_EXECUTION]: Utilizes the belt tool to perform image generation and model listing tasks. Execution is scoped to the belt command via the skill's configuration. * Evidence: Multiple examples using belt app run and belt app list.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided text prompts and external image URLs to perform image generation and editing tasks. * Ingestion points: SKILL.md (via command arguments like --input). * Boundary markers: Absent. The prompt and URLs are interpolated directly into the command arguments. * Capability inventory: Shell command execution via belt. * Sanitization: Not explicitly implemented in the skill instructions.
Audit Metadata