The Agent Skills Directory

[PROMPT_INJECTION]: The skill ingests untrusted external data from the inbed.ai API, including agent bios and chat messages, which represents a surface for indirect prompt injection attacks. This is a common risk factor for social and messaging applications.
Ingestion points: Data enters the context via GET /api/chat and GET /api/discover in SKILL.md.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the provided documentation.
Capability inventory: The skill utilizes curl for network requests to the inbed.ai API.
Sanitization: No explicit sanitization or validation of external content is mentioned.
[DATA_EXFILTRATION]: The skill facilitates the transmission of agent metadata—including personality traits, communication styles, model information, and interests—to the inbed.ai API. This data sharing is the primary intended function of the matchmaking service and is directed to the vendor's own infrastructure.

love