The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local Python script draw.py to interface with image generation models.
[DATA_EXFILTRATION]: The script draw.py reads sensitive files including .env and ~/.openai.env to retrieve the OPENAI_API_KEY. While necessary for the skill's function, this involves programmatic access to files that may contain other sensitive environment variables.
[EXTERNAL_DOWNLOADS]: The skill communicates with api.openai.com to send prompts and retrieve generated image data. This is an interaction with a well-known service required for the skill's primary operation.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through its processing of user-influenced arguments.
Ingestion points: Untrusted data is ingested via the prompt and --name arguments in draw.py.
Boundary markers: No delimiters or instructions are used to sanitize or isolate the provided input strings.
Capability inventory: The script performs directory creation and file writing operations on the local file system.
Sanitization: The --name parameter lacks validation, creating a surface for path traversal attacks if a malicious string is provided to redirect file output.

draw