og-image-design

Pass

Audited by Gen Agent Trust Hub on Sep 21, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill directs users to install external dependencies such as belt-sh/cli via npx and references installation instructions hosted on a third-party GitHub repository (github.com/inference-sh).
  • [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted user data and interpolates it into prompts for external AI and HTML-to-image services. * Ingestion points: The html parameter in infsh/html-to-image commands and the prompt parameter in falai/flux-dev-lora commands are derived from user-controlled content. * Boundary markers: There are no delimiters or explicit instructions provided to the agent to treat the interpolated content as data rather than instructions. * Capability inventory: The skill possesses the ability to execute shell commands via the belt CLI to interact with remote APIs. * Sanitization: The skill lacks validation, escaping, or sanitization mechanisms for the user input before it is passed to the shell environment.
  • [COMMAND_EXECUTION]: The skill uses the belt CLI to execute various applications (html-to-image, flux-dev-lora, search-assistant) which involve network communication and external processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 21, 2026, 08:14 PM
Security Audit — agent-trust-hub — og-image-design