from evaluation.framework_rag_benchmark import run_framework_benchmark


def test_real_langgraph_stategraph_rag_suite():
    result = run_framework_benchmark()
    assert result["clean"]["passed"] == result["clean"]["n_cases"] == 8
    assert result["conflict"]["passed"] == result["conflict"]["n_cases"] == 18
    assert {"retrieve", "assess", "generate", "abstain"}.issubset(result["stategraph_nodes"])