feat: enhance streaming implementation with proper tool call support

- Fix critical OpenAI client access bug: use sync_client instead of client
- Reduce code duplication by consolidating completion creation logic
- Improve test script: remove unused imports and properly test generator consumption
- Verify comprehensive tool call handling in streaming mode is working correctly
- All automated review feedback addressed while maintaining backward compatibility

Co-authored-by: Mervin Praison <MervinPraison@users.noreply.github.com>

Author: github-actions[bot]

src/praisonai-agents/praisonaiagents/agent/agent.py

@@ -2083,23 +2083,16 @@ def _start_stream(self, prompt: str, **kwargs) -> Generator[str, None, None]:
                     formatted_tools = self._format_tools_for_completion(tool_param)
 
                     # Create streaming completion directly without display function
+                    completion_args = {
+                        "model": self.llm,
+                        "messages": messages,
+                        "temperature": kwargs.get('temperature', 0.2),
+                        "stream": True
+                    }
                     if formatted_tools:
-                        # With tools - need to handle tool calls
-                        completion = self._openai_client.client.chat.completions.create(
-                            model=self.llm,
-                            messages=messages,
-                            temperature=kwargs.get('temperature', 0.2),
-                            tools=formatted_tools,
-                            stream=True
-                        )
-                    else:
-                        # Simple text completion
-                        completion = self._openai_client.client.chat.completions.create(
-                            model=self.llm,
-                            messages=messages,
-                            temperature=kwargs.get('temperature', 0.2),
-                            stream=True
-                        )
+                        completion_args["tools"] = formatted_tools
+                    
+                    completion = self._openai_client.sync_client.chat.completions.create(**completion_args)
 
                     # Stream the response chunks without display
                     response_text = ""

test_streaming_display_fix.py

@@ -6,7 +6,7 @@
 
 import sys
 import os
-import time
+import collections.abc
 
 # Add the praisonai-agents source to Python path
 sys.path.insert(0, os.path.join(os.path.dirname(__file__), 'src', 'praisonai-agents'))
@@ -28,16 +28,18 @@
     print("✅ Agent created successfully with stream=True")
     print(f"📊 Agent stream attribute: {agent.stream}")
 
-    # Test 2: Check start method behavior
+    # Test 2: Check start method behavior and exception on consumption
+    result = agent.start("Hello, test streaming")
+    assert isinstance(result, collections.abc.Generator), "Agent.start() should return a generator for streaming"
+    print("✅ Agent.start() returned a generator (streaming enabled)")
+
     try:
-        # This should use _start_stream method
-        result = agent.start("Hello, test streaming")
-        if hasattr(result, '__iter__') and hasattr(result, '__next__'):
-            print("✅ Agent.start() returned a generator (streaming enabled)")
-        else:
-            print("❌ Agent.start() did not return a generator")
+        # Consume the generator to trigger the API call, which should fail for a mock model.
+        list(result)
+        # If we get here, the test has failed because an exception was expected.
+        print("❌ FAILED: Expected an exception with mock model, but none was raised.")
     except Exception as e:
-        print(f"⚠️  Expected exception with mock model: {e}")
+        print(f"✅ SUCCESS: Caught expected exception with mock model: {e}")
         print("✅ Streaming path was triggered (exception expected with mock model)")
 
     # Test 3: Verify the streaming method exists and is callable