apple-reminders
Pass
Audited by Gen Agent Trust Hub on Aug 14, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses a native, on-device command
apple-remindersto interact with the EventKit framework. Analysis of the skill instructions and CLI reference confirms this tool is used only for local data management (reading, creating, updating, and deleting reminders). There are no shell injections or arbitrary command execution patterns detected; the skill provides strict guidelines on how to format arguments for this specific tool. - [DATA_EXPOSURE]: The skill instructions explicitly warn the agent not to surface a reminder's notes field unless specifically requested, as they may contain sensitive medical, financial, or personal information. This is a privacy-preserving design pattern.
- [SAFE]: The static analysis hint regarding 'concealment' is a false positive. The skill instructs the agent to suppress raw JSON, IDs, and technical details from the user-facing response to improve clarity and professional output, not to hide malicious actions. All high-impact actions (writes/deletes) are explicitly required to be reported back to the user with the exact items affected to ensure reviewability.
Audit Metadata