Async/sync FHIR client for Python providing comprehensive API for CRUD operations over FHIR resources
Agent Success
Agent success rate when using this tile
68%
Improvement
Agent success rate improvement when using this tile compared to baseline
1.08x
Baseline
Agent success rate without this tile
63%
{
"context": "Evaluates how well the solution leverages fhir-py reference helpers to build, normalize, resolve, and operate on FHIR references rather than reimplementing HTTP or string handling. Emphasis is on using the Reference helper API (to_reference, to_resource, execute) appropriately for each step in the workflow.",
"type": "weighted_checklist",
"checklist": [
{
"name": "Reference creation",
"description": "Creates references through fhir-py helpers (e.g., client.reference(resource_type, resource_id, display=...)) instead of manually concatenating strings.",
"max_score": 20
},
{
"name": "Canonical export",
"description": "Normalizes inputs by calling Reference.to_reference() to emit canonical reference dicts, preserving display when provided.",
"max_score": 20
},
{
"name": "Resource resolution",
"description": "Uses Reference.to_resource() to hydrate references when a resource is not already provided, and avoids redundant fetches when the resource object is present.",
"max_score": 15
},
{
"name": "Operation execution",
"description": "Invokes operations via Reference.execute(operation, method=..., data/params=...) rather than manual HTTP calls, returning the operation result.",
"max_score": 25
},
{
"name": "Display handling",
"description": "Passes display text through fhir-py reference creation so it appears in the emitted Reference structure.",
"max_score": 10
},
{
"name": "Mixed input support",
"description": "Accepts strings, reference dicts, and resource dicts/objects by converting them through fhir-py Reference helpers instead of hand-rolled parsing.",
"max_score": 10
}
]
}evals
scenario-1
scenario-2
scenario-3
scenario-4
scenario-5
scenario-6
scenario-7
scenario-8
scenario-9
scenario-10