Unsupported filter string format on `mlclient.search_registered_models`

- **Package Name**: `azureml-mlflow`
- **Package Version**: `1.60.0.post1`
- **Operating System**: Linux
- **Python Version**: `3.12

**Describe the bug**
I am using MLFlow to query for registered models in AzureML.

When calling `mlclient.search_registered_models` I get the following error:
```
mlflow.exceptions.RestException: INVALID_PARAMETER_VALUE: Response: {'Error': {'Code': 'UserError', 'Severity': None, 'Message': "Currently unsupported filter string format (we currently do not support OR). Provided filter string: tag.`mlflow.prompt.is_prompt` != 'true'", 'MessageFormat': None, 'MessageParameters': None, 'ReferenceCode': None, 'DetailsUri': None, 'Target': None, 'Details': [], 'InnerError': None, 'DebugInfo': None, 'AdditionalInfo': None}, 'Correlation': {'operation': 'ac2c85060fa14b5f2d628e9dc95f8c03', 'request': '0051f7fc64471137'}, 'Environment': 'uksouth', 'Location': 'uksouth', 'Time': '2025-07-22T10:22:46.55723+00:00', 'ComponentName': 'mlflow', 'statusCode': 400, 'error_code': 'INVALID_PARAMETER_VALUE'}
```

**To Reproduce**
Steps to reproduce the behavior:
```python
from mlflow import MlflowClient

mlclient = MlflowClient()
versions = mlclient.search_registered_models("name='my-model-name'", order_by=["version"])
```

**Expected behavior**
Should return model versions.

**Additional context**
If I downgrade to `azureml-mlflow==1.58.0` and `mlflow-skinny==2.17.0` allows this to work as before.