fix: add flow_context field to CrossDatabaseState to preserve topology data between nodes

Without this field, LangGraph silently drops the topology data returned
by identify_flow() before verify_mechanism() can read it, making the
entire topological flow reasoning feature a no-op in production.

Also adds a regression test (test_flow_context_in_state) to ensure
the field is never accidentally removed.

Author: bhavyakeerthi3

src/agent/profiles/cross_database.py

@@ -31,6 +31,8 @@ class CrossDatabaseState(BaseState):
     uniprot_answer: str  # LLM-generated answer from UniProt
     uniprot_completeness: str  # LLM-assessed completeness of the UniProt answer
 
+    flow_context: str  # Topological flow data fetched by identify_flow() for mechanistic queries
+
 
 class CrossDatabaseGraphBuilder(BaseGraphBuilder):
     def __init__(

tests/test_flow_reasoning.py

@@ -101,3 +101,21 @@ async def test_decide_next_steps_mechanistic(builder):
     }
     decision = await builder.decide_next_steps(state)
     assert decision == "generate_final_response"
+
+
+def test_flow_context_in_state():
+    """
+    Regression test: flow_context must be declared in CrossDatabaseState.
+
+    Without this field, LangGraph silently drops the topology data returned
+    by identify_flow() before verify_mechanism() can read it — making the
+    entire topological reasoning feature a no-op.
+    """
+    from agent.profiles.cross_database import CrossDatabaseState
+    import typing
+
+    hints = typing.get_type_hints(CrossDatabaseState)
+    assert "flow_context" in hints, (
+        "CrossDatabaseState is missing the 'flow_context' field. "
+        "LangGraph will silently drop topology data between nodes without it."
+    )