The Agent Skills Directory

[SAFE]: The skill performs its documented purpose of setting up a wellness automation for the user without any malicious side effects.
[COMMAND_EXECUTION]: Uses the create_scheduled_task tool to establish a recurring automation for sending affirmations, which is consistent with the skill's description.
[DATA_EXFILTRATION]: Requests user contact information (SMS or email) solely for the delivery of the requested messages. No data is transmitted to unauthorized external services.
[PROMPT_INJECTION]: The skill allows the user to provide a list of affirmations, creating a surface for indirect instructions to be processed during the scheduled task execution.
Ingestion points: User-provided affirmation lists described in Step 1 of SKILL.md.
Boundary markers: None provided in the task template.
Capability inventory: Uses create_scheduled_task to generate and send content.
Sanitization: No specific sanitization or filtering of the affirmation content is implemented.

daily-affirmations