The Agent Skills Directory

[DYNAMIC_EXECUTION]: The scripts/check-reminder-logic.mjs script utilizes Node.js import() and module hooks to dynamically load TypeScript modules from the project being checked. This mechanism is designed to allow the skill's logic checker to run and verify the application's actual scheduling and reconciliation logic against test vectors.
[INDIRECT_PROMPT_INJECTION]: The skill defines a surface for processing untrusted data that could potentially contain malicious instructions. Ingestion points: Untrusted data enters the agent context through NotificationTap objects from the operating system (handled in use-reminder-taps.ts) and Reminder rows from the local database which include user-supplied labels. Boundary markers: While the skill emphasizes using structured data, it does not implement specific delimiters or 'ignore' instructions for the interpolated reminder labels in notification bodies. Capability inventory: The skill can schedule or cancel system notifications and trigger navigation to specific application routes based on reminder IDs. Sanitization: The skill mitigates risks by implementing the parseReminderTap function in reminder-tap.ts, which validates the schema and content of tap data, and by using internationalization helpers to isolate user-supplied text during notification generation.

local-notifications