ToTick Task Management

Problem

When an agent must turn natural language into a persistent task system, raw HTTP calls become brittle: device tokens, subscription quotas, and the mapping from AI-parsed cards to note creation fields are easy to miss. ToTick Task Management packages these concerns into an API skill, so agents can consistently use device-token, inspect subscription state, parse text, and save structured notes.

How It Works

• Natural-language parsing: call the AI parsing endpoint to split a sentence into todo, meeting, reminder, or note cards, with fields such as title, content, dueDate, reminderAt, and action items.
• Save and retrieve: map cards to the create-note request, then use GET /api/v1/notes, widget-digest, search, and stats endpoints to read todo summaries, overdue items, and today's dashboard.
• Operational loop: support completing, archiving, restoring, snoozing or dismissing reminders, searching participants, and managing tags or projects, turning conversational intents like “note this” or “mark done” into concrete API calls.
• Presentation constraints: output Markdown cards, relative time labels, and done/total progress while avoiding exposure of sensitive data such as email or device token.

Boundaries

The skill depends on the ToTick backend and the user's device token. When the token expires, quota is exhausted, parsing fails, or note access is denied, surface the error message instead of swallowing it silently. Time uses yyyy-MM-dd HH:mm:ss, AI parsing has monthly quotas, and many endpoints require device-token authentication.