ai-image-gen
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided descriptions to generate image prompts, which represents an attack surface for indirect prompt injection.
- Ingestion points: User instructions are interpolated into command-line arguments for the
text2imgandimg2imgsubcommands inSKILL.md. - Boundary markers: The instructions do not specify any delimiters or safety warnings to ignore instructions embedded within user-supplied text.
- Capability inventory: The skill has the capability to perform network requests to external API providers and write files to the
outputs/directory. - Sanitization: There is no evidence of input sanitization or filtering in the instructional logic; the skill relies on the underlying LLM's safety filters.
- [COMMAND_EXECUTION]: The skill invokes a local Python script
skills/shared/scripts/ai_image.pyvia the shell to handle configuration checks and image generation tasks. - [EXTERNAL_DOWNLOADS]: The skill facilitates the download of generated image files from remote API providers (OpenAI or Apimart) to the local project directory.
Audit Metadata