How scheduling works
When you create a schedule, Dialog registers your research prompt with the scheduling engine and begins executing it at the intervals you define. Each execution runs in the same persistent workspace environment as your chat sessions, so scheduled tasks have access to all your workspace files. Here’s the flow:- You create a schedule with a prompt, cron expression, and delivery preferences
- The scheduling engine triggers your task at the specified times
- Dialog runs your research prompt just like an interactive chat session
- Results are delivered to your chosen channels — web inbox, Telegram, or both
Schedule configuration
When creating a schedule, you’ll provide:- Name — a label to identify this schedule in your dashboard
- Cron expression — defines when the task runs (e.g.,
0 9 * * 1-5for weekday mornings at 9 AM) - Timezone — defaults to UTC, but you can set it to your local timezone
- Prompt — the research question or instructions, up to 2,000 characters
- Delivery channels — where results go, defaults to web inbox
Task status lifecycle
Every scheduled task has a status that tells you where it stands.
If a task fails three times in a row, Dialog automatically pauses it to prevent wasted credits. When you resume a failed task, the failure counter resets and the next run time is recalculated from the current time.
Delivery
Scheduled task results are delivered through your configured channels. You can choose:- Web inbox only — results appear in the app, ready when you are
- Telegram only — results pushed directly to your Telegram chat
- Both — the same results delivered to web inbox and Telegram simultaneously

