The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a fallback Python script (scripts/image_gen.py) designed to interact with the OpenAI API for image tasks. This script is intended for explicit user-requested batch or advanced operations and handles file system writes for generated assets within defined project paths.
[EXTERNAL_DOWNLOADS]: The skill documentation and scripts reference the openai and pillow Python packages. These are well-known, industry-standard libraries used for API communication and image processing, respectively.
[INDIRECT_PROMPT_INJECTION]: As the skill's primary function is to transform user-supplied text prompts into images, it naturally possesses an attack surface for indirect prompt injection.
Ingestion points: User-provided prompts are received through both the built-in image_gen tool and the scripts/image_gen.py CLI.
Boundary markers: The skill uses a structured prompt schema (e.g., Use case, Asset type, Primary request) to help distinguish user intent, although this is not a strict security boundary.
Capability inventory: The skill code can write generated image data to the local filesystem and make outbound network requests to the OpenAI API.
Sanitization: The skill relies on the underlying LLM and the external API's moderation features, as it does not perform manual sanitization of the prompt strings before processing.

imagegen