ai-product-photography

Pass

Audited by Gen Agent Trust Hub on Aug 31, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill makes extensive use of the belt CLI tool via Bash commands to execute AI image generation tasks. The access is restricted via the allowed-tools frontmatter to Bash(belt *) which follows the principle of least privilege.
  • [EXTERNAL_DOWNLOADS]: The skill references the installation of the belt-sh/cli package via npx and provides a link to an external installation guide on GitHub (https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md). These resources are directly related to the vendor's own infrastructure for the service provided.
  • [INDIRECT_PROMPT_INJECTION]: The skill template constructs shell commands by interpolating user-provided prompts into JSON strings for the CLI input. This creates a surface where malicious external data could attempt to influence the command execution.
  • Ingestion points: Prompt content for product descriptions and variables like PRODUCT in the batch generation example (SKILL.md).
  • Boundary markers: None explicitly defined in the provided Bash templates.
  • Capability inventory: Execution of the belt command via Bash subprocesses.
  • Sanitization: No explicit sanitization logic is provided in the documentation, relying on the agent's default handling of string interpolation.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 31, 2026, 11:00 PM
Security Audit — agent-trust-hub — ai-product-photography