refactor: rename and add is error check

Author: norman-le

src/uipath_langchain/agent/react/agent.py

@@ -81,22 +81,22 @@ def create_agent(
     llm_tools: list[BaseTool] = [*agent_tools, *flow_control_tools]
 
     # Derive client-side tool schemas from tools for input validation in the init node.
-    cs_tools: dict[str, ClientSideToolInfo] | None = None
+    conversational_client_side_tools: dict[str, ClientSideToolInfo] | None = None
     if config.is_conversational:
-        cs_tools = {}
+        conversational_client_side_tools = {}
         for t in agent_tools:
             meta = getattr(t, "metadata", None) or {}
             if meta.get(IS_CONVERSATIONAL_CLIENT_SIDE_TOOL):
-                cs_tools[t.name] = {
+                conversational_client_side_tools[t.name] = {
                     "input_schema": t.args_schema.model_json_schema()
                     if hasattr(t, "args_schema") and t.args_schema
                     else None,
                     "output_schema": meta.get("output_schema"),
                 }
-        cs_tools = cs_tools or None
+        conversational_client_side_tools = conversational_client_side_tools or None
 
     init_node = create_init_node(
-        messages, input_schema, config.is_conversational, cs_tools
+        messages, input_schema, config.is_conversational, conversational_client_side_tools
     )
 
     tool_nodes = create_tool_node(agent_tools)

src/uipath_langchain/agent/tools/client_side_tool.py

@@ -98,21 +98,25 @@ async def wait_for_client_execution() -> dict[str, Any]:
                 }
 
             result = await wait_for_client_execution()
-            return result.get("output", result) if isinstance(result, dict) else result
+            return result if isinstance(result, dict) else {"output": result}
 
         result = await execute_tool()
 
-        if isinstance(result, dict):
+        is_error = result.get("isError", False)
+        output = result.get("output", result)
+
+        if isinstance(output, dict):
             try:
-                content = json.dumps(result)
+                content = json.dumps(output)
             except TypeError:
-                content = str(result)
+                content = str(output)
         else:
-            content = str(result) if result is not None else ""
+            content = str(output) if output is not None else ""
 
         return ToolMessage(
             content=content,
             tool_call_id=tool_call_id,
+            status="error" if is_error else "success",
             response_metadata={IS_CONVERSATIONAL_CLIENT_SIDE_TOOL: True},
         )