The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill interpolates user-controlled inputs directly into the prompts used for image and video generation.
Ingestion points: User inputs baby_costume, dialogue, and baby_description in SKILL.md.
Boundary markers: No boundary markers or instructions are provided to isolate user data from agent instructions.
Capability inventory: The skill utilizes the muapi CLI and curl for network operations.
Sanitization: No sanitization of user-provided content is performed before interpolation.
[DATA_EXFILTRATION]: The skill performs network operations to the external domain api.muapi.ai which is not a whitelisted or well-known service.
Evidence: The agent is instructed to use curl to send POST requests to https://api.muapi.ai/api/v1/<endpoint>.
Risk: The skill transmits user-provided content and an authentication token (MUAPI_API_KEY) to this external service to perform its primary function.

muapi-talking-baby-video