Align run-history helper pagination with result listings.

Add offset support to list_run_history and extend tests/contracts so runtime pagination semantics stay consistent across normalized helper APIs.

Author: eddietejeda

CONTRACT.md

@@ -45,7 +45,7 @@ Adapters should import from `hotdata_runtime` and treat this surface as the stab
 - `query_runs()` returns the query-runs API wrapper for adapter history views.
 - `results()` returns the results API wrapper for adapter result pickers.
 - `list_recent_results(...)` returns normalized `ResultSummary` entries.
-- `list_run_history(...)` returns normalized `RunHistoryItem` entries.
+- `list_run_history(limit=..., offset=...)` returns normalized `RunHistoryItem` entries.
 - `list_qualified_table_names(...)` returns sorted fully qualified table names.
 - `columns_for_qualified(qualified, connection_id=...)` resolves table columns, and
   adapters should pass `connection_id` when known.

hotdata_runtime/client.py

@@ -152,8 +152,9 @@ def list_run_history(
         self,
         *,
         limit: int = 20,
+        offset: int = 0,
     ) -> list[RunHistoryItem]:
-        listing = self.query_runs().list_query_runs(limit=limit)
+        listing = self.query_runs().list_query_runs(limit=limit, offset=offset)
         return [
             RunHistoryItem(
                 query_run_id=r.id,

tests/test_client.py

@@ -221,11 +221,17 @@ def test_list_run_history_returns_normalized_items():
     )
 
     class FakeRunsApi:
-        def list_query_runs(self, *, limit: int):
+        def __init__(self):
+            self.kwargs = None
+
+        def list_query_runs(self, *, limit: int, offset: int):
+            self.kwargs = {"limit": limit, "offset": offset}
             return listing
 
-    with patch.object(client, "query_runs", return_value=FakeRunsApi()):
-        out = client.list_run_history(limit=5)
+    fake_runs = FakeRunsApi()
+    with patch.object(client, "query_runs", return_value=fake_runs):
+        out = client.list_run_history(limit=5, offset=3)
     assert [r.query_run_id for r in out] == ["run_1"]
     assert out[0].execution_time_ms == 7
     assert out[0].to_dict()["result_id"] == "res_1"
+    assert fake_runs.kwargs == {"limit": 5, "offset": 3}