It would be really useful to support recurring agent jobs and commands, so users can schedule an agent command to run automatically at a specific time and cadence.
Problem
Right now agent commands are primarily manual. That makes it harder to automate repetitive workflows such as daily project maintenance, routine checks/summaries, scheduled code review/codegen tasks, or repo housekeeping.
Proposed solution
Add a way to configure scheduled recurring jobs for agents, for example:
- run every day at 09:00
- run every weekday at 18:00
- run every Monday morning
- optionally support cron-style schedules for advanced users
A simple version could allow a user to:
- choose an agent command or script to run
- choose a schedule
- optionally choose the working directory / repo
- review logs or results from each run
Expected behavior
- recurring schedules persist across app restarts
- users can view, pause, edit, and delete scheduled jobs
- each run has visible logs / status
- failures are surfaced clearly
- safeguards exist for overlapping runs or long-running jobs
Nice-to-haves
- timezone selection
- notifications on success/failure
- retry policy
- per-job environment / repo selection
- support for both simple schedules and advanced cron syntax
Why this would help
This would make the tool much more useful for ongoing automation workflows, not just one-off agent interactions.
It would be really useful to support recurring agent jobs and commands, so users can schedule an agent command to run automatically at a specific time and cadence.
Problem
Right now agent commands are primarily manual. That makes it harder to automate repetitive workflows such as daily project maintenance, routine checks/summaries, scheduled code review/codegen tasks, or repo housekeeping.
Proposed solution
Add a way to configure scheduled recurring jobs for agents, for example:
A simple version could allow a user to:
Expected behavior
Nice-to-haves
Why this would help
This would make the tool much more useful for ongoing automation workflows, not just one-off agent interactions.