The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external image data provided via URLs, which introduces a surface for instructions hidden within the data itself.
Ingestion points: The image_urls parameter in the API request examples in SKILL.md allows the agent to fetch and process arbitrary external images.
Boundary markers: The provided prompts lack explicit instructions to the AI model to ignore or separate any text-based instructions that might be present within the processed images (e.g., OCR-based injections).
Capability inventory: The skill enables network operations via curl to the eachsense-agent.core.eachlabs.run API endpoint and processes complex generation responses.
Sanitization: There is no evidence of validation or sanitization mechanisms for the content retrieved from the external image URLs before it is processed by the AI models.

age-transformation