The Agent Skills Directory

[SAFE]: The skill interacts with the vendor's API using platform-provided environment variables ($AUTH_TOKEN and $API_URL). This is a standard and safe way to handle authentication.
[COMMAND_EXECUTION]: Uses curl to send structured JSON data to the scheduling service. The usage is restricted to the intended API operations (create, list, update, delete).
[DATA_EXFILTRATION]: User-provided prompts are sent to the cloud backend to be stored for scheduling. This is the primary function of the skill and does not involve exfiltrating sensitive local files or environment data.
[PROMPT_INJECTION]: The skill has a surface for indirect prompt injection because it stores user-supplied strings for future execution.
Ingestion points: prompt and title fields in SKILL.md.
Boundary markers: Not specified in the instructions for the API payload.
Capability inventory: Network access via curl to the vendor's API.
Sanitization: Not explicitly implemented in the skill's logic, relying on the platform's underlying security controls.

cloud-schedule