The Agent Skills Directory

[COMMAND_EXECUTION]: The skill documentation provides numerous curl command templates to demonstrate how to interact with the each::sense API endpoint at https://sense.eachlabs.run/chat. These examples include the use of session IDs for iterative editing and various processing modes.
[EXTERNAL_DOWNLOADS]: The skill facilitates the processing of external images by sending user-provided URLs to the vendor's API. The provided documentation in references/SSE-EVENTS.md also indicates that generated outputs are hosted on the vendor's storage infrastructure at storage.eachlabs.ai.
[PROMPT_INJECTION]: The skill represents an indirect prompt injection surface as it processes external data (images and user messages) through an AI model.
Ingestion points: External data enters the context via the image_urls and message fields in the API requests defined in SKILL.md.
Boundary markers: There are no explicit boundary markers or instructions to ignore potential commands embedded within the input data mentioned in the documentation.
Capability inventory: The skill's core capability is executing image manipulation requests against an external AI service.
Sanitization: No specific sanitization or validation steps for input data are described within the provided skill files.

eye-color-changer