Frequently Asked Questions
What happens if my token quota runs out?
Agent jobs will fail. You can see the failure reason on any job from AI Agents > Jobs, and check your remaining quota from Settings > Tokens. See Tokens & AI Usage.
Which task types show up on the Backlog?
Task, Bug, Improvement, and Research. Epics and Stories don't — they're organizational units of work, not deliverables, and live on the Product Delivery and Road Map boards instead.
Can I create a Research task through AI generation?
Yes — select Research as the task type when generating with the PO agent, same as Task, Bug, Improvement, Epic, or Story. See Generating Tasks.
Can I assign a Research task to the Coder or Tester agent?
No. A Research task's output is findings (investigations, spikes), not code, so it can't be assigned to the Coder or Tester agent — only Task, Bug, and Improvement tasks can.
What's the difference between the changelog and the release log?
Both are generated by the PO agent from the Releases board. The changelog is technical documentation, useful for tracking and debugging production issues. The release log is written in end-user-friendly language, meant for announcing updates to your project's users.
Do dedicated worker instances consume token quota?
No. Workloads that run inside a dedicated instance don't consume tokens and get exclusive priority on that instance. See Workers.
Can the Tester agent run against a production environment?
No — never provide it with production credentials. The Tester should only create and run automations against a validation/test environment that you control and keep up to date with the version you want tested.
Who do I contact for help?
General support: support@caesell.ai. For questions about dedicated worker instances or pricing: sales@caesell.ai.