The Agent Skills Directory

[COMMAND_EXECUTION]: The skill defines patterns for the agent to execute curl commands in the shell to interact with the Venice AI API endpoints (e.g., https://api.venice.ai/api/v1/image/edit).
[EXTERNAL_DOWNLOADS]: The skill allows the agent to fetch and process images from external HTTPS URLs provided by the user via the image or image_url parameters.
[DATA_EXFILTRATION]: The skill transmits user-supplied prompts and image data (in base64 format or as URLs) to the Venice AI API. These operations are essential for the skill's functionality and target the official service domain.
[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it interpolates untrusted user data into API requests.
Ingestion points: Untrusted data enters the agent context through the prompt, image, and image_url fields across all image editing endpoints defined in SKILL.md.
Boundary markers: The skill instructions do not prescribe the use of delimiters or 'ignore' instructions to mitigate the risk of the agent obeying malicious commands embedded within user prompts or image metadata.
Capability inventory: The skill has the capability to perform network requests to api.venice.ai and read/process image data provided by the user.
Sanitization: There is no evidence of input validation, escaping, or filtering of the user-provided content before it is processed or sent to the external API.

venice-image-edit