feat(retry): consume uipath-llm-client 1.17.0 and raise default max retries to 5

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "uipath-langchain"
-version = "0.14.9"
+version = "0.14.10"
 description = "Python SDK that enables developers to build and deploy LangGraph agents to the UiPath Cloud Platform"
 readme = { file = "README.md", content-type = "text/markdown" }
 requires-python = ">=3.11"
@@ -9,7 +9,7 @@ dependencies = [
     "uipath-core>=0.5.29, <0.6.0",
     "uipath-platform>=0.2.10, <0.3.0",
     "uipath-runtime>=0.12.1, <0.13.0",
-    "uipath-llm-client>=1.16.3, <1.17.0",
+    "uipath-llm-client>=1.17.0, <1.18.0",
     "langgraph>=1.1.8, <2.0.0",
     "langchain-core>=1.2.27, <2.0.0",
     "langgraph-checkpoint-sqlite>=3.0.3, <4.0.0",
@@ -26,7 +26,7 @@ dependencies = [
     "pillow>=12.1.1",
     "rdflib>=7.0.0, <8.0.0",
     "a2a-sdk>=0.2.0,<1.0.0",
-    "uipath-langchain-client[openai]>=1.16.1,<1.17.0",
+    "uipath-langchain-client[openai]>=1.17.0,<1.18.0",
 ]
 
 classifiers = [
@@ -43,21 +43,21 @@ maintainers = [
 
 [project.optional-dependencies]
 anthropic = [
-    "uipath-langchain-client[anthropic]>=1.16.1,<1.17.0",
+    "uipath-langchain-client[anthropic]>=1.17.0,<1.18.0",
 ]
 vertex = [
-    "uipath-langchain-client[google]>=1.16.1,<1.17.0",
-    "uipath-langchain-client[vertexai]>=1.16.1,<1.17.0",
+    "uipath-langchain-client[google]>=1.17.0,<1.18.0",
+    "uipath-langchain-client[vertexai]>=1.17.0,<1.18.0",
 ]
 bedrock = [
-    "uipath-langchain-client[bedrock]>=1.16.1,<1.17.0",
+    "uipath-langchain-client[bedrock]>=1.17.0,<1.18.0",
     "boto3-stubs>=1.41.4",
 ]
 fireworks = [
-    "uipath-langchain-client[fireworks]>=1.16.1,<1.17.0",
+    "uipath-langchain-client[fireworks]>=1.17.0,<1.18.0",
 ]
 all = [
-    "uipath-langchain-client[all]>=1.16.1,<1.17.0",
+    "uipath-langchain-client[all]>=1.17.0,<1.18.0",
 ]
 
 [project.entry-points."uipath.middlewares"]

src/uipath_langchain/chat/chat_model_factory.py

@@ -59,7 +59,9 @@ def on_llm_start(
 DEFAULT_TIMEOUT_SECONDS: Final[float] = 895.0
 DEFAULT_MAX_TOKENS: Final[int] = 1000
 DEFAULT_TEMPERATURE: Final[float] = 0.0
-DEFAULT_MAX_RETRIES: Final[int] = 3
+# 5 total attempts (initial request + up to 4 retries) for parity with the
+# legacy in-repo retryers (BedrockRetryer / VertexRetryer) and uipath-llm-client defaults.
+DEFAULT_MAX_RETRIES: Final[int] = 5
 
 
 def get_chat_model(
@@ -99,7 +101,8 @@ def get_chat_model(
             forward an explicit unset value (lets the underlying client apply
             its own default or use no limit).
         timeout: Request timeout in seconds. Defaults to 895 seconds.
-        max_retries: Max retry count. Defaults to 3.
+        max_retries: Maximum number of attempts (ignored when
+            use_new_llm_clients=False). Defaults to 5.
         callbacks: LangChain callbacks (handlers or a manager) attached to the
             returned chat model. Accepts ``list[BaseCallbackHandler]`` or a
             ``BaseCallbackManager``. Forwarded only when explicitly set.

tests/chat/test_chat_model_factory.py

@@ -465,3 +465,31 @@ def test_gemini_vertex_forwards_temperature(self, mocker):
 
         _, kwargs = mock_cls.call_args
         assert kwargs.get("temperature") == 0.5
+
+
+class TestDefaultMaxRetriesForwarding:
+    """Pins ``DEFAULT_MAX_RETRIES`` to 5 (parity with the legacy in-repo
+    retryers) and asserts the default reaches the ``uipath_langchain_client``
+    factory on the new-clients path, so the old default cannot be silently
+    reintroduced."""
+
+    def test_default_max_retries_is_5(self):
+        from uipath_langchain.chat.chat_model_factory import DEFAULT_MAX_RETRIES
+
+        assert DEFAULT_MAX_RETRIES == 5
+
+    def test_default_max_retries_forwarded_to_new_client_factory(self, mocker):
+        from uipath_langchain.chat import chat_model_factory as outer_factory
+
+        mock_factory = mocker.patch(
+            "uipath_langchain.chat.chat_model_factory.get_chat_model_factory"
+        )
+
+        outer_factory.get_chat_model(
+            "gpt-test",
+            agenthub_config="cfg",
+            use_new_llm_clients=True,
+        )
+
+        _, kwargs = mock_factory.call_args
+        assert kwargs["max_retries"] == 5