Scheduled Jobs
You can ask the AI to schedule tasks that run automatically on a recurring basis. Whether you want a daily news summary, a weekly report, or a regular check on something that matters to you, jobs let the AI work for you in the background without you having to remember to ask.
Creating a job
Just ask the AI in plain language. For example:
“Check the latest AI news every morning at 9am and summarize it for me.”
“Every Monday, remind me to review my open tasks.”
“Run a web search for competitor pricing once a week and send me a summary.”
The AI will set up the job and confirm the schedule. The new job appears on the Scheduled Jobs page, which you can open from your avatar menu (bottom of the left rail) or directly at /jobs.
The Jobs page
All your scheduled jobs are listed there, filterable by All, Active, and Paused. Each job shows:
- The job name
- The status (active or paused)
- The schedule (how often it runs)
Select a job to see its details: the full schedule (type, timezone, next and last run), the trigger prompt, and its recent runs. The Open chat button jumps to the conversation where the job posts its results.
Managing jobs
Pausing and resuming
If you want to temporarily stop a job without removing it, click Pause. The job won’t run again until you resume it. Click Resume to re-enable it.
Triggering a job now
To run a job immediately without waiting for its next scheduled time, click Run now. This is useful for testing a new job or getting a fresh result outside the normal schedule.
Changing a schedule
To change when or how often a job runs, ask the AI in the job’s conversation, for example “run this daily at 7am instead”. The AI updates the job for you.
Stopping a job
To remove a job for good, click Stop. The job stops running and disappears from your job list.
Job history and results
Each job’s detail view lists its recent runs with status, start time, and duration. The actual output of a run (summaries, messages, generated content) lands in the job’s conversation; use Open chat to read it.
If a run fails, the list shows an error status with a short error message. You can try running it again manually or ask the AI to adjust the job if the error is due to an invalid configuration.