Migrate agent policies to a new fleet server

Hello everyone! i have a test cluster with fleet server where there is an elastic agent with synthetics (heartbeat) policy. as this is a test cluster i decided to make a single policy "fleet+heartbeat policy" and assign it to a single agent. but now i have a task to separate this policy into two and migrate them to prod cluster, so it has to be "fleet policy" and "heartbeat policy" assigned to two separate agents
my questions are:

  1. how to correctly separate policy into two? i tried to copy it, remove fleet integration, but already have some issues, for example - to all heartbeat monitors names were added a postfix (copy); i can't delete these copied policies and etc
  2. how to import policies to a new cluster?
  3. can i migrate agents with current policies to a new cluster and will they continue working with these policies or they will be assigned to new one, that exist in a new cluster? will migrated policies be added to a list of policies on a new cluster?
    btw i already red a topic Redirect Elastic Agent to new Fleet Server, there wasn't any correct answer

You can edit a policy and change the name while duplicating it.

I'm not sure about the heartbeat because I do not use, but you will need to edit the policy and change the names of the monitor.

There is a documentation about it, but in my opinion is not good: Migrate Fleet-managed Elastic Agents from one cluster to another | Fleet and Elastic Agent Guide [8.17] | Elastic

It assumes that you will restore a snapshot of your current cluster on another cluster, and all the examples are from the Elastic Cloud offering.

It does not says how you can migrate just the fleet configurations, what indices or features needs to be migrated etc.

I'm in a similar process and I'm recreating everything in the new cluster, policies, integrations etc.

You need to re-enroll the Agents in the new cluster in the new policies.

In my opinion you will spend less time if you just recreate everything in the new cluster, the migration process is not well documented, there is no examples for on-prem clusters and there is no example that does not envolves restoring a snapshot of the entire cluster.