The Agent Skills Directory

[DATA_EXFILTRATION]: The skill documents and facilitates network requests to https://sense.eachlabs.run/chat. This domain belongs to the skill's vendor (Each Labs) and is required for the image generation service. As the traffic is directed to the vendor's own infrastructure, this is considered a functional requirement.
[PROMPT_INJECTION]: The documentation includes instructions to set enable_safety_checker to false. While this terminology overlaps with common safety bypass patterns, it is a documented technical parameter of the Each Labs API specifically intended to allow the generation of adult and nude art as part of the skill's primary purpose.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by processing untrusted user input and external data.
Ingestion points: User-provided text prompts in the messages object and external image references in the image_urls array (as seen in SKILL.md and the client examples in references/SSE-EVENTS.md).
Boundary markers: No explicit delimiters or instructions to ignore embedded commands (e.g., in image metadata or prompt text) are provided in the payload construction templates.
Capability inventory: The skill transmits data via network requests (using curl, fetch, or requests) to the vendor's API endpoint.
Sanitization: The documentation notes server-side enforcement of content policies and provides error handling for policy violations (content_policy_violation).

NSFW Image Generation