[rohangujarathi]

Issue #, if available:

Description of changes:

Adds a pipeline step that imports content into a SageMaker hub.

• Session.import_hub_content(...) (sagemaker-core): builds an ImportHubContent request and routes it through _intercept_create_request, so under a PipelineSession it returns step arguments instead of calling the service. hub_content_document can be a string or a dict; a dict is serialized when the request is sent.
• HubContentStep (sagemaker-mlops): a ConfigurableRetryStep that accepts step_args from import_hub_content only, supports cache_config and retry_policies, and exposes properties shaped like DescribeHubContentResponse. A dict document is encoded in arguments the same way JobStep encodes JobConfigDocument, so pipeline variables inside it are resolved at execution time.
• StepTypeEnum.HUB_CONTENT = "HubContent" and the sagemaker.mlops.workflow export.

Example:

pipeline_session = PipelineSession()
step_args = pipeline_session.import_hub_content(
    hub_name="my-hub",
    hub_content_name="my-model",
    hub_content_type="Model",
    document_schema_version="2.4.0",
    hub_content_document={...},
)
step = HubContentStep(name="ImportModel", step_args=step_args)
# step.properties.HubContentArn, step.properties.HubContentStatus, ...

When hub_content_version is omitted, the service assigns the next version on each execution.

The HubContent step type is being rolled out in SageMaker Pipelines; please hold the merge until it is available, otherwise CreatePipeline rejects the definition and the new integ tests fail.

Testing done:

• Unit tests in sagemaker-mlops/tests/unit/workflow/test_hub_content_step.py: capture without a service call, request shape, unset optional fields, dict and string documents, pipeline-variable document encoding, step request with cache and retry config, producer validation, properties, pipeline definition, plain Session call.
• Unit tests in TestImportHubContent (sagemaker-core/tests/unit/helper/test_session_helper.py): required fields, dict serialization, all optional fields and tags, empty optional values, service error propagation.
• test_job_step.py enum member set updated; existing test_inference_lineage_steps.py, test_job_step.py and test_session_helper.py pass.
• Integ tests in sagemaker-mlops/tests/integ/workflow/test_hub_content_step.py: an import with a parameterized document feeding a ConditionStep on HubContentStatus and HubContentType, unpinned reruns importing a new version, and a missing hub failing the step. All 3 pass.

Merge Checklist

General

• I have read the CONTRIBUTING doc
• I certify that the changes I am introducing will be backward compatible, and I have discussed concerns about this, if any, with the Python SDK team
• I used the commit message format described in CONTRIBUTING
• I have passed the region in to all S3 and STS clients that I've initialized as part of this change.
• I have updated any necessary documentation, including READMEs and API docs (if appropriate)

Tests

• I have added tests that prove my fix is effective or that my feature works (if appropriate)
• I have added unit and/or integration tests as appropriate to ensure backward compatibility of the changes
• I have checked that my tests are not configured for a specific region or account (if appropriate)
• I have used unique_name_from_base to create resource names in integ tests (if appropriate)
• If adding any dependency in requirements.txt files, I have spell checked and ensured they exist in PyPi

By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.

[github-actions]

🤖 AI code review

This PR adds Session.import_hub_content(...) in sagemaker-core and a HubContentStep pipeline step in sagemaker-mlops, plus a StepTypeEnum.HUB_CONTENT value and workflow export. The implementation closely follows existing patterns: import_hub_content routes through _intercept_create_request like other producers, dict documents are serialized via the shared convert_job_config_document_to_string (so pipeline variables become a Join), step_args is validated with validate_step_args_input, and properties is shaped from the botocore DescribeHubContentResponse shape. I verified the imports (json, format_tags, _append_project_tags, RequestType, _JobStepArguments, Properties), the base-class signatures, and that arguments reads a shallow copy of step_args.args without mutating the captured request. Unit and integ tests are comprehensive. No correctness, security, or backward-compatibility issues found; the implementation is clean.

Note: the PR body asks maintainers to hold the merge until the HubContent step type is available service-side (otherwise CreatePipeline/integ tests fail). That is operational context for reviewers, not a code issue.

_{Reviewed commit 909574f. Automated review; verify before acting.}

[github-actions]

🤖 AI code review

This PR adds Session.import_hub_content() (sagemaker-core) and HubContentStep (sagemaker-mlops), a ConfigurableRetryStep that imports content into a SageMaker hub within a pipeline. The producer routes through _intercept_create_request so it returns step args under a PipelineSession, and a dict HubContentDocument is encoded the same way JobStep encodes JobConfigDocument so pipeline variables resolve at execution time. The implementation closely follows the established JobStep/TrainingStep patterns, the StepTypeEnum addition and workflow export are correct, and the unit tests cover request shape, dict/string documents, pipeline-variable encoding, producer validation, properties, and the plain-Session path. I found no correctness or security issues; the code is consistent with existing conventions. Note: the PR body asks reviewers to hold the merge until the HubContent step type is rolled out in Pipelines (otherwise the integ tests fail) — that is a release-coordination item, not a code concern.

_{Reviewed commit e5d26ec. Automated review; verify before acting.}

[github-actions]

🤖 AI code review

This PR adds Session.import_hub_content() in sagemaker-core (routing through _intercept_create_request so it returns step args under a PipelineSession, serializing a dict HubContentDocument at request time) and a HubContentStep in sagemaker-mlops, plus StepTypeEnum.HUB_CONTENT and the workflow export. The implementation closely mirrors the established JobStep/JobConfigDocument pattern: the arguments property shallow-copies the captured request and encodes a dict document via convert_job_config_document_to_string for pipeline-variable resolution, and to_request folds in cache config. Imports (json, format_tags, _append_project_tags, validate_step_args_input) are all present at the base commit, tag/empty-value handling matches existing methods, and unit + integ tests cover capture, serialization, optional fields, producer validation, properties, and pipeline definition. I found no correctness, security, or backward-compatibility issues. Note: the PR body asks reviewers to hold the merge until the HubContent step type is rolled out server-side — a process note, not something that changes the code review.

_{Reviewed commit e9882e0. Automated review; verify before acting.}