Identify sudden spikes in cost, be it per-object (container, pod, project, cluster) or aggregated.

Let user define what "unusual" means and in what scope:

• For some customers a 10% increase might be worrying, for others a 30% increase might be fine
• A 30% spending increase in a development cluster might be fine (eg. cluster was mostly idle or even stopped, now it's being used), while a 10% increase in a production cluster might be worrying.