p-image

Pass

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of the belt CLI and associated components from official repositories hosted by the service provider on GitHub.
  • [COMMAND_EXECUTION]: Utilizes the belt CLI tool to perform authentication and execute inference tasks with various AI models.
  • [INDIRECT_PROMPT_INJECTION]: Ingests user-provided text prompts and external image URLs as inputs for model execution. The risk is mitigated by passing these inputs through a structured JSON schema to the underlying CLI tool.
  • Ingestion points: input JSON object in belt app run commands (SKILL.md)
  • Boundary markers: Inputs are encapsulated within a JSON string structure
  • Capability inventory: Bash(belt *) tool for CLI interactions (SKILL.md)
  • Sanitization: Input processing and validation are delegated to the belt CLI utility
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 2, 2026, 02:18 PM
Security Audit — agent-trust-hub — p-image