{
  "context": "Evaluates whether the solution builds and executes the Observation search using fhir-py query utilities. Checks correct use of chaining, comparison formatting, reference conversion, boolean encoding, and raw parameter passthrough.",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "Chained identifier",
      "description": "Patient constraint is built with fhir-py's chained search support (e.g., SQ(subject__Patient__identifier=...) or SearchSet.search with double-underscore chaining) instead of manual query strings.",
      "max_score": 25
    },
    {
      "name": "Datetime bound",
      "description": "Uses fhir-py's comparison helpers (such as SQ(effective__ge=recorded_after) or search(date__ge=recorded_after)) to format the lower-bound datetime with the proper ge prefix rather than hand-crafting the timestamp string.",
      "max_score": 20
    },
    {
      "name": "Reference formatting",
      "description": "Formats performer filtering through fhir-py's reference helpers (e.g., client.reference('Practitioner', performer_id) or to_reference()) so the query sends a Practitioner/{id} reference instead of concatenated strings.",
      "max_score": 20
    },
    {
      "name": "Boolean handling",
      "description": "Passes boolean status flags directly into SearchSet.search so fhir-py performs the true/false encoding, avoiding manual string literals for the status filter.",
      "max_score": 15
    },
    {
      "name": "Raw passthrough",
      "description": "Adds any extra expression via fhirpy.base.searchset.Raw (or equivalent Raw wrapper) to append the literal query segment without transformation.",
      "max_score": 20
    }
  ]
}