Remove null values from CDS Hook API responses

[jenniferjiangkells]

CDS Hook API responses currently include null values for all optional fields, making responses verbose and harder to read. This is especially noticeable in example scripts and documentation.

Example current output:

{
  "cards": [{
    "summary": "Sepsis Risk: 0.1",
    "indicator": "info",
    "source": {
      "label": "ML prediction",
      "url": null,
      "icon": null,
      "topic": null
    },
    "uuid": null,
    "detail": "No action required.",
    "suggestions": null,
    "selectionBehavior": null,
    "overrideReasons": null,
    "links": null
  }],
  "systemActions": null
}

Desired output:

{
  "cards": [{
    "summary": "Sepsis Risk: 0.1",
    "indicator": "info",
    "source": {
      "label": "ML prediction"
    },
    "detail": "No action required."
  }]
}

Files to check:

• scripts/predict_example.py (confirmed issue)
• Any file calling .model_dump() on CDS models
• healthchain/gateway/cds/ (CDS service implementations)
• Example scripts in cookbook/

Goal:
Configure Pydantic serialization to exclude None values from API responses.

Acceptance Criteria:

• Find all locations where CDS models are serialized (search for .model_dump())
• Add exclude_none=True parameter to serialization calls
• Verify responses are cleaner in examples (test manually)
• Ensure all tests still pass
• Response validation should still work correctly

Implementation:

Card(...).model_dump(exclude_none=True)

Resources:

• Pydantic serialization docs
• healthchain/models/responses/cdsresponse.py for model definitions
• Test with: python scripts/predict_example.py and check http://localhost:8000/docs

[esther-anierobi]

Hi @jenniferjiangkells , I would like to work on this. Please assign to me.

[jenniferjiangkells]

@esther-anierobi You're welcome to start working on it!

[sathwikshetty33]

Hi @jenniferjiangkells , I would like to work on this. Please assign to me, if not yet resolved.
Thank you

[esther-anierobi]

Hi @sathwikshetty33 , I've already made a PR and some corrections waiting for review. Please, kindly look for another issue to contribute to. Thank you.