dbt test --store-failures only create *audit db in one node

### Describe the bug

The dbt test, when applied with --store-failures, only creates the *audit db in one node, ignoring the dbt profiles config `cluster: '{cluster_name}'`.

### Steps to reproduce
1. Create a dbt profile with the cluster config
2. Run dbt test --store-failures.. using the config at one
3. Inspect your clickhouse cluster and check if the *audit db has been created on all nodes

### Expected behaviour

The *audit db should be created on all nodes give the cluster config is specified in the profile

### Code examples, such as models or profile settings
```yaml
data_platform_clickhouse_dbt:
target: prod_raw
outputs:
prod_raw:
type: clickhouse
schema: "{{ 'prod_metrics_' + var('pr_id') if var('pr_id', '') != '' else 'prod_metrics' }}" # Append 'metrics_' only if 'pr_id' is set
host: demo-host
port: 8443
user: USER
password: PWD
secure: true
verify: false
cluster: '{cluster}'
```

`dbt test --profiles-dir=./profiles/prod --vars "{'pr_id': '${{ env.PR_ID }}'}" --store-failures --select state:modified --state ./state --defer`

### Configuration
#### Environment
+ dbt-adapters==1.16.3
+ dbt-common==1.27.1
+ dbt-core==1.10.7
+ dbt-extractor==0.6.0
+ dbt-protos==1.0.348
+ dbt-semantic-interfaces==0.9.0
+ clickhouse-driver==0.2.9
+ dbt-clickhouse==1.8.9
+ dbt-core==1.8.9

#### ClickHouse server
* ClickHouse Server version: 25.5.7

Start by reproducing the `dbt test --store-failures` command with the shown profile and a configured `cluster`, then trace how dbt-clickhouse creates the failure-storage audit database. Compare the database creation behavior with the profile's cluster setting. Done means the `*audit` database is created on every ClickHouse cluster node when that setting is present.

Written by the indexing model from the issue text.