fix: select nodepool id when describe cluster attach scripts

[Vacant2333]

What type of PR is this?

/kind bug

What this PR does / why we need it:

We encountered a situation where the DescribeClusterAttachScriptsRequest will use the default nodepool when the NodepoolID is not provided. However, in some cases, the default nodepool may have been deleted, but DescribeClusterAttachScriptsRequest will still use the CRI configuration of the default nodepool. In an ACK cluster with version 1.24, if the default nodepool was originally created with Docker Runtime but was later deleted and replaced with a new nodepool using Containerd Runtime, this could cause nodes to fail to register to the cluster. This is because Docker Runtime only supports clusters of version 1.22 and below now.

Which issue(s) this PR fixes:

Fixes #

Special notes for your reviewer:

Does this PR introduce a user-facing change?

To call `DescribeClusterAttachScriptsRequest`, Karpenter need to provide the NodePoolID now, preferably selecting the default NodePool. If you are using Karpenter deployed with minimal permissions after account transfer, you need to grant your RAM User the permission cs:DescribeClusterNodePools. This will resolve the issue where nodes created by older clusters may fail to register with the cluster in some cases.

[jwcesign]

/lgtm