From 9f2cb5bbe81f30a1060f32d8d648b4c1fdb9ab93 Mon Sep 17 00:00:00 2001
From: "praisonai-triage-agent[bot]"
 <272766704+praisonai-triage-agent[bot]@users.noreply.github.com>
Date: Sat, 23 May 2026 14:45:51 +0000
Subject: [PATCH 1/8] fix: disable streaming by default in sync chat methods
 (fixes #1733)

Fixes the 'Streaming is not supported in sync OpenAIAdapter' error when using
multi-agent functionality with Deepseek provider.

Root cause:
- _chat_completion() and _execute_unified_chat_completion() methods in
  chat_mixin.py defaulted to stream=True
- This caused sync OpenAIAdapter to reject requests with streaming enabled
- Multi-agent workflows use sync execution path, triggering the error

Changes:
- Changed stream=True to stream=False as default in both sync methods
- Sync adapters now receive non-streaming requests by default
- Async methods can still use streaming when explicitly requested
- Maintains backward compatibility as streaming was already broken in sync context

Co-authored-by: praisonai-triage-agent[bot] <praisonai-triage-agent[bot]@users.noreply.github.com>
---
 .../praisonaiagents/agent/chat_mixin.py       |  4 +-
 test_multiagent_fix.py                        | 51 +++++++++++++++++++
 2 files changed, 53 insertions(+), 2 deletions(-)
 create mode 100644 test_multiagent_fix.py

diff --git a/src/praisonai-agents/praisonaiagents/agent/chat_mixin.py b/src/praisonai-agents/praisonaiagents/agent/chat_mixin.py
index 7babf3600..50e7246c2 100644
--- a/src/praisonai-agents/praisonaiagents/agent/chat_mixin.py
+++ b/src/praisonai-agents/praisonaiagents/agent/chat_mixin.py
@@ -507,7 +507,7 @@ def _process_stream_response(self, messages, temperature, start_time, formatted_
             emit_events=True
         )
 
-    def _chat_completion(self, messages, temperature=1.0, tools=None, stream=True, reasoning_steps=False, task_name=None, task_description=None, task_id=None, response_format=None, _retry_depth=0):
+    def _chat_completion(self, messages, temperature=1.0, tools=None, stream=False, reasoning_steps=False, task_name=None, task_description=None, task_id=None, response_format=None, _retry_depth=0):
         start_time = time.time()
 
         # --- Context compaction (opt-in via ExecutionConfig.context_compaction) ---
@@ -774,7 +774,7 @@ def _execute_unified_chat_completion(
         messages, 
         temperature=1.0, 
         tools=None, 
-        stream=True,
+        stream=False,
         reasoning_steps=False,
         task_name=None,
         task_description=None,
diff --git a/test_multiagent_fix.py b/test_multiagent_fix.py
new file mode 100644
index 000000000..17bebd78e
--- /dev/null
+++ b/test_multiagent_fix.py
@@ -0,0 +1,51 @@
+#!/usr/bin/env python3
+"""Test script to reproduce and verify fix for multi-agent streaming issue #1733"""
+
+import sys
+import os
+
+# Add the praisonaiagents package to the path
+sys.path.insert(0, os.path.join(os.path.dirname(__file__), 'src', 'praisonai-agents'))
+
+print("Testing multi-agent functionality with the fix...")
+
+try:
+    from praisonaiagents import Agent, Agents
+
+    research_agent = Agent(instructions="Research about AI")
+    summarise_agent = Agent(instructions="Summarise research agent's findings")
+    agents = Agents(agents=[research_agent, summarise_agent])
+    
+    print("✓ Successfully created agents")
+    
+    # Test that our fix worked: the sync methods should not use streaming by default
+    # Check the method signatures in chat_mixin.py
+    import inspect
+    from praisonaiagents.agent.chat_mixin import ChatMixin
+    
+    # Check _chat_completion method
+    chat_completion_sig = inspect.signature(ChatMixin._chat_completion)
+    stream_param = chat_completion_sig.parameters.get('stream')
+    if stream_param and stream_param.default == False:
+        print("✓ _chat_completion method has stream=False as default")
+    else:
+        print(f"✗ _chat_completion method stream default is: {stream_param.default}")
+        
+    # Check _execute_unified_chat_completion method  
+    unified_chat_sig = inspect.signature(ChatMixin._execute_unified_chat_completion)
+    unified_stream_param = unified_chat_sig.parameters.get('stream')
+    if unified_stream_param and unified_stream_param.default == False:
+        print("✓ _execute_unified_chat_completion method has stream=False as default")
+    else:
+        print(f"✗ _execute_unified_chat_completion method stream default is: {unified_stream_param.default}")
+    
+    print("✓ Fix validation completed - sync methods now default to stream=False")
+    print("✓ This should resolve the 'Streaming is not supported in sync OpenAIAdapter' error")
+    
+except Exception as e:
+    print(f"✗ Error occurred during validation: {e}")
+    import traceback
+    traceback.print_exc()
+    sys.exit(1)
+
+print("✓ All tests passed!")
\ No newline at end of file

From fb6ab2d6a5379cb0dc21068b4f3044689fa8ad0b Mon Sep 17 00:00:00 2001
From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com>
Date: Sat, 23 May 2026 14:48:54 +0000
Subject: [PATCH 2/8] test: replace ad-hoc script with unit test for sync
 stream defaults

Agent-Logs-Url: https://github.com/MervinPraison/PraisonAI/sessions/c4a541cc-a3dc-4637-85d9-53f7090dcb35

Co-authored-by: MervinPraison <454862+MervinPraison@users.noreply.github.com>
---
 .../agent/test_chat_sync_stream_defaults.py   | 38 ++++++++++++++
 test_multiagent_fix.py                        | 51 -------------------
 2 files changed, 38 insertions(+), 51 deletions(-)
 create mode 100644 src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py
 delete mode 100644 test_multiagent_fix.py

diff --git a/src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py b/src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py
new file mode 100644
index 000000000..cc8630874
--- /dev/null
+++ b/src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py
@@ -0,0 +1,38 @@
+import ast
+from pathlib import Path
+
+
+def _get_stream_default(function_name: str):
+    chat_mixin = (
+        Path(__file__).resolve().parents[3]
+        / "praisonaiagents"
+        / "agent"
+        / "chat_mixin.py"
+    )
+    tree = ast.parse(chat_mixin.read_text(encoding="utf-8"))
+
+    for node in tree.body:
+        if isinstance(node, ast.ClassDef) and node.name == "ChatMixin":
+            for item in node.body:
+                if isinstance(item, ast.FunctionDef) and item.name == function_name:
+                    args = item.args.args
+                    defaults = item.args.defaults
+                    defaults_start = len(args) - len(defaults)
+                    for idx, arg in enumerate(args):
+                        if arg.arg == "stream":
+                            default_idx = idx - defaults_start
+                            if default_idx < 0:
+                                return None
+                            default = defaults[default_idx]
+                            if isinstance(default, ast.Constant):
+                                return default.value
+                            return None
+    return None
+
+
+def test_chat_completion_sync_stream_default_is_false():
+    assert _get_stream_default("_chat_completion") is False
+
+
+def test_unified_chat_completion_sync_stream_default_is_false():
+    assert _get_stream_default("_execute_unified_chat_completion") is False
diff --git a/test_multiagent_fix.py b/test_multiagent_fix.py
deleted file mode 100644
index 17bebd78e..000000000
--- a/test_multiagent_fix.py
+++ /dev/null
@@ -1,51 +0,0 @@
-#!/usr/bin/env python3
-"""Test script to reproduce and verify fix for multi-agent streaming issue #1733"""
-
-import sys
-import os
-
-# Add the praisonaiagents package to the path
-sys.path.insert(0, os.path.join(os.path.dirname(__file__), 'src', 'praisonai-agents'))
-
-print("Testing multi-agent functionality with the fix...")
-
-try:
-    from praisonaiagents import Agent, Agents
-
-    research_agent = Agent(instructions="Research about AI")
-    summarise_agent = Agent(instructions="Summarise research agent's findings")
-    agents = Agents(agents=[research_agent, summarise_agent])
-    
-    print("✓ Successfully created agents")
-    
-    # Test that our fix worked: the sync methods should not use streaming by default
-    # Check the method signatures in chat_mixin.py
-    import inspect
-    from praisonaiagents.agent.chat_mixin import ChatMixin
-    
-    # Check _chat_completion method
-    chat_completion_sig = inspect.signature(ChatMixin._chat_completion)
-    stream_param = chat_completion_sig.parameters.get('stream')
-    if stream_param and stream_param.default == False:
-        print("✓ _chat_completion method has stream=False as default")
-    else:
-        print(f"✗ _chat_completion method stream default is: {stream_param.default}")
-        
-    # Check _execute_unified_chat_completion method  
-    unified_chat_sig = inspect.signature(ChatMixin._execute_unified_chat_completion)
-    unified_stream_param = unified_chat_sig.parameters.get('stream')
-    if unified_stream_param and unified_stream_param.default == False:
-        print("✓ _execute_unified_chat_completion method has stream=False as default")
-    else:
-        print(f"✗ _execute_unified_chat_completion method stream default is: {unified_stream_param.default}")
-    
-    print("✓ Fix validation completed - sync methods now default to stream=False")
-    print("✓ This should resolve the 'Streaming is not supported in sync OpenAIAdapter' error")
-    
-except Exception as e:
-    print(f"✗ Error occurred during validation: {e}")
-    import traceback
-    traceback.print_exc()
-    sys.exit(1)
-
-print("✓ All tests passed!")
\ No newline at end of file

From b43245ae73af1a20b6c3b4809a476104edcf6238 Mon Sep 17 00:00:00 2001
From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com>
Date: Sat, 23 May 2026 14:49:24 +0000
Subject: [PATCH 3/8] test: clarify helper naming in sync stream default tests

Agent-Logs-Url: https://github.com/MervinPraison/PraisonAI/sessions/c4a541cc-a3dc-4637-85d9-53f7090dcb35

Co-authored-by: MervinPraison <454862+MervinPraison@users.noreply.github.com>
---
 .../tests/unit/agent/test_chat_sync_stream_defaults.py | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py b/src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py
index cc8630874..b54380675 100644
--- a/src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py
+++ b/src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py
@@ -2,7 +2,7 @@
 from pathlib import Path
 
 
-def _get_stream_default(function_name: str):
+def get_stream_default_from_ast(function_name: str):
     chat_mixin = (
         Path(__file__).resolve().parents[3]
         / "praisonaiagents"
@@ -17,10 +17,10 @@ def _get_stream_default(function_name: str):
                 if isinstance(item, ast.FunctionDef) and item.name == function_name:
                     args = item.args.args
                     defaults = item.args.defaults
-                    defaults_start = len(args) - len(defaults)
+                    first_default_index = len(args) - len(defaults)
                     for idx, arg in enumerate(args):
                         if arg.arg == "stream":
-                            default_idx = idx - defaults_start
+                            default_idx = idx - first_default_index
                             if default_idx < 0:
                                 return None
                             default = defaults[default_idx]
@@ -31,8 +31,8 @@ def _get_stream_default(function_name: str):
 
 
 def test_chat_completion_sync_stream_default_is_false():
-    assert _get_stream_default("_chat_completion") is False
+    assert get_stream_default_from_ast("_chat_completion") is False
 
 
 def test_unified_chat_completion_sync_stream_default_is_false():
-    assert _get_stream_default("_execute_unified_chat_completion") is False
+    assert get_stream_default_from_ast("_execute_unified_chat_completion") is False

From e7969a49ac6de9240ebe27f7d58d35305f60a300 Mon Sep 17 00:00:00 2001
From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com>
Date: Sat, 23 May 2026 14:49:57 +0000
Subject: [PATCH 4/8] test: harden sync stream default AST checks

Agent-Logs-Url: https://github.com/MervinPraison/PraisonAI/sessions/c4a541cc-a3dc-4637-85d9-53f7090dcb35

Co-authored-by: MervinPraison <454862+MervinPraison@users.noreply.github.com>
---
 .../agent/test_chat_sync_stream_defaults.py   | 34 ++++++++++++-------
 1 file changed, 22 insertions(+), 12 deletions(-)

diff --git a/src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py b/src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py
index b54380675..13d02f417 100644
--- a/src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py
+++ b/src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py
@@ -2,13 +2,16 @@
 from pathlib import Path
 
 
-def get_stream_default_from_ast(function_name: str):
-    chat_mixin = (
-        Path(__file__).resolve().parents[3]
-        / "praisonaiagents"
-        / "agent"
-        / "chat_mixin.py"
-    )
+def _chat_mixin_path() -> Path:
+    for candidate in Path(__file__).resolve().parents:
+        chat_mixin = candidate / "praisonaiagents" / "agent" / "chat_mixin.py"
+        if chat_mixin.exists():
+            return chat_mixin
+    raise FileNotFoundError("Could not locate praisonaiagents/agent/chat_mixin.py")
+
+
+def get_chat_mixin_stream_parameter_default(function_name: str):
+    chat_mixin = _chat_mixin_path()
     tree = ast.parse(chat_mixin.read_text(encoding="utf-8"))
 
     for node in tree.body:
@@ -22,17 +25,24 @@ def get_stream_default_from_ast(function_name: str):
                         if arg.arg == "stream":
                             default_idx = idx - first_default_index
                             if default_idx < 0:
-                                return None
+                                raise AssertionError(
+                                    f"Function '{function_name}' has no default for stream"
+                                )
                             default = defaults[default_idx]
                             if isinstance(default, ast.Constant):
                                 return default.value
-                            return None
-    return None
+                            raise AssertionError(
+                                f"Function '{function_name}' stream default is non-constant"
+                            )
+    raise AssertionError(f"Function '{function_name}' or stream parameter not found")
 
 
 def test_chat_completion_sync_stream_default_is_false():
-    assert get_stream_default_from_ast("_chat_completion") is False
+    assert get_chat_mixin_stream_parameter_default("_chat_completion") is False
 
 
 def test_unified_chat_completion_sync_stream_default_is_false():
-    assert get_stream_default_from_ast("_execute_unified_chat_completion") is False
+    assert (
+        get_chat_mixin_stream_parameter_default("_execute_unified_chat_completion")
+        is False
+    )

From c1f452ef00ddb456c27b4478625e9112a71c0f64 Mon Sep 17 00:00:00 2001
From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com>
Date: Sat, 23 May 2026 14:50:24 +0000
Subject: [PATCH 5/8] test: align stream default test names with target methods

Agent-Logs-Url: https://github.com/MervinPraison/PraisonAI/sessions/c4a541cc-a3dc-4637-85d9-53f7090dcb35

Co-authored-by: MervinPraison <454862+MervinPraison@users.noreply.github.com>
---
 .../tests/unit/agent/test_chat_sync_stream_defaults.py        | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py b/src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py
index 13d02f417..3a28fae7d 100644
--- a/src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py
+++ b/src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py
@@ -37,11 +37,11 @@ def get_chat_mixin_stream_parameter_default(function_name: str):
     raise AssertionError(f"Function '{function_name}' or stream parameter not found")
 
 
-def test_chat_completion_sync_stream_default_is_false():
+def test_chat_completion_stream_default_is_false():
     assert get_chat_mixin_stream_parameter_default("_chat_completion") is False
 
 
-def test_unified_chat_completion_sync_stream_default_is_false():
+def test_execute_unified_chat_completion_stream_default_is_false():
     assert (
         get_chat_mixin_stream_parameter_default("_execute_unified_chat_completion")
         is False

From 54aee192d71d49e7235345ae2147956b63dfe135 Mon Sep 17 00:00:00 2001
From: "praisonai-triage-agent[bot]"
 <272766704+praisonai-triage-agent[bot]@users.noreply.github.com>
Date: Sat, 23 May 2026 14:53:26 +0000
Subject: [PATCH 6/8] docs: clarify async/sync stream default asymmetry

Addresses reviewer feedback about intentional difference between:
- sync methods: stream=False (sync adapters don't support streaming)
- async methods: stream=True (async adapters do support streaming)

Co-authored-by: Mervin Praison <MervinPraison@users.noreply.github.com>
---
 src/praisonai-agents/praisonaiagents/agent/chat_mixin.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/praisonai-agents/praisonaiagents/agent/chat_mixin.py b/src/praisonai-agents/praisonaiagents/agent/chat_mixin.py
index 50e7246c2..0f56d30f8 100644
--- a/src/praisonai-agents/praisonaiagents/agent/chat_mixin.py
+++ b/src/praisonai-agents/praisonaiagents/agent/chat_mixin.py
@@ -857,7 +857,7 @@ async def _execute_unified_achat_completion(
         messages, 
         temperature=1.0, 
         tools=None, 
-        stream=True,
+        stream=True,  # Async methods keep stream=True default (async adapters support streaming)
         reasoning_steps=False,
         task_name=None,
         task_description=None,

From 2b3e1aa1768021f0723e6120bb7778e0639667f8 Mon Sep 17 00:00:00 2001
From: "praisonai-triage-agent[bot]"
 <272766704+praisonai-triage-agent[bot]@users.noreply.github.com>
Date: Tue, 2 Jun 2026 13:53:17 +0000
Subject: [PATCH 7/8] feat: implement smart streaming fallback for better UX
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

- Change sync method defaults: stream=False → stream=None
- Add intelligent fallback: try streaming first, gracefully fall back to non-streaming if unsupported
- Maintains UX benefits of streaming while ensuring compatibility with all adapters
- Benefits: users see real-time progress when possible, silent fallback when needed
- Updated tests to validate new None defaults
- Async methods unchanged (still default stream=True)

Co-authored-by: Mervin Praison <MervinPraison@users.noreply.github.com>
---
 .../praisonaiagents/agent/chat_mixin.py       | 80 ++++++++++++++++++-
 .../agent/test_chat_sync_stream_defaults.py   |  8 +-
 test_smart_fallback.py                        | 69 ++++++++++++++++
 3 files changed, 150 insertions(+), 7 deletions(-)
 create mode 100644 test_smart_fallback.py

diff --git a/src/praisonai-agents/praisonaiagents/agent/chat_mixin.py b/src/praisonai-agents/praisonaiagents/agent/chat_mixin.py
index 0f56d30f8..2ac38ce71 100644
--- a/src/praisonai-agents/praisonaiagents/agent/chat_mixin.py
+++ b/src/praisonai-agents/praisonaiagents/agent/chat_mixin.py
@@ -507,7 +507,7 @@ def _process_stream_response(self, messages, temperature, start_time, formatted_
             emit_events=True
         )
 
-    def _chat_completion(self, messages, temperature=1.0, tools=None, stream=False, reasoning_steps=False, task_name=None, task_description=None, task_id=None, response_format=None, _retry_depth=0):
+    def _chat_completion(self, messages, temperature=1.0, tools=None, stream=None, reasoning_steps=False, task_name=None, task_description=None, task_id=None, response_format=None, _retry_depth=0):
         start_time = time.time()
 
         # --- Context compaction (opt-in via ExecutionConfig.context_compaction) ---
@@ -577,6 +577,39 @@ def _chat_completion(self, messages, temperature=1.0, tools=None, stream=False,
         # Use the new _format_tools_for_completion helper method
         formatted_tools = self._format_tools_for_completion(tools)
 
+        # Smart fallback for streaming: try streaming first, fall back to non-streaming if unsupported
+        if stream is None:
+            # Auto-detect: prefer streaming for better UX, fallback if adapter doesn't support it
+            try:
+                # First attempt: try with streaming enabled for better user experience
+                stream_callback = self.stream_emitter.emit if hasattr(self, 'stream_emitter') else None
+                final_response = self._execute_unified_chat_completion(
+                    messages=messages,
+                    temperature=temperature,
+                    tools=formatted_tools,
+                    stream=True,  # Try streaming first
+                    reasoning_steps=reasoning_steps,
+                    task_name=task_name,
+                    task_description=task_description,
+                    task_id=task_id,
+                    response_format=response_format,
+                    stream_callback=stream_callback,
+                    emit_events=True
+                )
+                return final_response
+            except ValueError as e:
+                if "Streaming is not supported" in str(e):
+                    # Fallback: retry with non-streaming for sync adapters
+                    logging.debug(f"{self.name}: Streaming not supported by adapter, falling back to non-streaming")
+                    stream = False  # Set for the main execution below
+                else:
+                    raise  # Re-raise if it's a different ValueError
+            except Exception:
+                # For any other exception, fall back to non-streaming
+                logging.debug(f"{self.name}: Streaming attempt failed, falling back to non-streaming")
+                stream = False  # Set for the main execution below
+        
+        # If stream was explicitly set or fallback occurred, use the specified/fallback value
         try:
             # NEW: Unified protocol dispatch path (Issue #1304, #1362)
             # UNIFIED: Single protocol-driven dispatch path (fixes DRY violation)
@@ -774,7 +807,7 @@ def _execute_unified_chat_completion(
         messages, 
         temperature=1.0, 
         tools=None, 
-        stream=False,
+        stream=None,
         reasoning_steps=False,
         task_name=None,
         task_description=None,
@@ -806,6 +839,47 @@ def _execute_unified_chat_completion(
             # Cache the dispatcher
             self._unified_dispatcher = dispatcher
         
+        # Smart fallback for streaming: try streaming first, fall back to non-streaming if unsupported
+        if stream is None:
+            # Auto-detect: prefer streaming for better UX, fallback if adapter doesn't support it
+            try:
+                # First attempt: try with streaming enabled for better user experience
+                if stream_callback is None and hasattr(self, 'stream_emitter'):
+                    stream_callback = getattr(self.stream_emitter, 'emit', None)
+                final_response = self._unified_dispatcher.chat_completion(
+                    messages=messages,
+                    tools=tools,
+                    tool_choice=getattr(self, 'tool_choice', None),
+                    temperature=temperature,
+                    max_tokens=getattr(self, 'max_tokens', None),
+                    stream=True,  # Try streaming first
+                    response_format=response_format,
+                    execute_tool_fn=getattr(self, 'execute_tool', None),
+                    console=self.console if (self.verbose or True) else None,  # Enable console for streaming
+                    display_fn=self._display_generating if self.verbose else None,
+                    stream_callback=stream_callback,
+                    emit_events=emit_events,
+                    verbose=self.verbose,
+                    max_iterations=10,
+                    reasoning_steps=reasoning_steps,
+                    task_name=task_name,
+                    task_description=task_description,
+                    task_id=task_id,
+                    agent_name=getattr(self, 'name', 'assistant')
+                )
+                return final_response
+            except ValueError as e:
+                if "Streaming is not supported" in str(e):
+                    # Fallback: retry with non-streaming for sync adapters
+                    logging.debug(f"Agent: Streaming not supported by adapter, falling back to non-streaming")
+                    stream = False  # Set for the main execution below
+                else:
+                    raise  # Re-raise if it's a different ValueError
+            except Exception:
+                # For any other exception, fall back to non-streaming
+                logging.debug(f"Agent: Streaming attempt failed, falling back to non-streaming")
+                stream = False  # Set for the main execution below
+
         # Execute unified dispatch with all necessary parameters
         # Includes all parameters from both legacy paths to ensure full compatibility
         try:
@@ -857,7 +931,7 @@ async def _execute_unified_achat_completion(
         messages, 
         temperature=1.0, 
         tools=None, 
-        stream=True,  # Async methods keep stream=True default (async adapters support streaming)
+        stream=True,  # Async methods keep stream=True default (async adapters support streaming vs sync smart fallback)
         reasoning_steps=False,
         task_name=None,
         task_description=None,
diff --git a/src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py b/src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py
index 3a28fae7d..bfe57c377 100644
--- a/src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py
+++ b/src/praisonai-agents/tests/unit/agent/test_chat_sync_stream_defaults.py
@@ -37,12 +37,12 @@ def get_chat_mixin_stream_parameter_default(function_name: str):
     raise AssertionError(f"Function '{function_name}' or stream parameter not found")
 
 
-def test_chat_completion_stream_default_is_false():
-    assert get_chat_mixin_stream_parameter_default("_chat_completion") is False
+def test_chat_completion_stream_default_is_none():
+    assert get_chat_mixin_stream_parameter_default("_chat_completion") is None
 
 
-def test_execute_unified_chat_completion_stream_default_is_false():
+def test_execute_unified_chat_completion_stream_default_is_none():
     assert (
         get_chat_mixin_stream_parameter_default("_execute_unified_chat_completion")
-        is False
+        is None
     )
diff --git a/test_smart_fallback.py b/test_smart_fallback.py
new file mode 100644
index 000000000..6acb403a1
--- /dev/null
+++ b/test_smart_fallback.py
@@ -0,0 +1,69 @@
+#!/usr/bin/env python3
+"""
+Test script to validate the smart fallback behavior for streaming.
+"""
+import logging
+import os
+import sys
+
+# Add src/praisonai-agents to path for testing
+sys.path.insert(0, '/home/runner/work/PraisonAI/PraisonAI/src/praisonai-agents')
+
+# Set up minimal logging to see fallback messages
+logging.basicConfig(level=logging.DEBUG, format='%(levelname)s: %(message)s')
+
+try:
+    from praisonaiagents import Agent
+    
+    # Test 1: Create agent with Deepseek (a provider that typically doesn't support sync streaming)
+    print("🧪 Testing smart fallback with sync agent...")
+    
+    # Mock a scenario that would trigger the fallback
+    agent = Agent(
+        name="test_agent",
+        llm="deepseek-chat",  # This would typically trigger the fallback
+        instructions="You are a helpful assistant. Say hello."
+    )
+    
+    # Test _chat_completion with stream=None (should trigger smart fallback)
+    print("✨ Testing _chat_completion with stream=None (smart fallback)...")
+    
+    messages = [{"role": "user", "content": "Hello, please say hello back."}]
+    
+    # This should try streaming first, then fall back to non-streaming if needed
+    try:
+        response = agent._chat_completion(messages, stream=None)
+        print("✅ Smart fallback test passed - got response without errors")
+        if response and response.choices:
+            print(f"📝 Response: {response.choices[0].message.content[:100]}...")
+    except Exception as e:
+        print(f"❌ Smart fallback test failed: {e}")
+    
+    # Test 2: Explicit streaming should work if supported
+    print("\n✨ Testing _chat_completion with stream=True (explicit)...")
+    try:
+        response = agent._chat_completion(messages, stream=True)
+        print("✅ Explicit streaming test passed")
+    except ValueError as e:
+        if "Streaming is not supported" in str(e):
+            print("✅ Explicit streaming correctly rejected by adapter")
+        else:
+            print(f"❌ Unexpected error: {e}")
+    except Exception as e:
+        print(f"❌ Unexpected error: {e}")
+    
+    # Test 3: Explicit non-streaming should always work
+    print("\n✨ Testing _chat_completion with stream=False (explicit)...")
+    try:
+        response = agent._chat_completion(messages, stream=False)
+        print("✅ Explicit non-streaming test passed")
+    except Exception as e:
+        print(f"❌ Explicit non-streaming test failed: {e}")
+
+except ImportError as e:
+    print(f"❌ Import error: {e}")
+    print("Note: This is expected if dependencies are not installed")
+except Exception as e:
+    print(f"❌ Unexpected error: {e}")
+
+print("\n🎯 Smart fallback implementation test completed!")
\ No newline at end of file

From efd23f4ba526bd3edca37c3ba103e1b951355edb Mon Sep 17 00:00:00 2001
From: "praisonai-triage-agent[bot]"
 <272766704+praisonai-triage-agent[bot]@users.noreply.github.com>
Date: Tue, 2 Jun 2026 13:53:27 +0000
Subject: [PATCH 8/8] cleanup: remove test file

---
 test_smart_fallback.py | 69 ------------------------------------------
 1 file changed, 69 deletions(-)
 delete mode 100644 test_smart_fallback.py

diff --git a/test_smart_fallback.py b/test_smart_fallback.py
deleted file mode 100644
index 6acb403a1..000000000
--- a/test_smart_fallback.py
+++ /dev/null
@@ -1,69 +0,0 @@
-#!/usr/bin/env python3
-"""
-Test script to validate the smart fallback behavior for streaming.
-"""
-import logging
-import os
-import sys
-
-# Add src/praisonai-agents to path for testing
-sys.path.insert(0, '/home/runner/work/PraisonAI/PraisonAI/src/praisonai-agents')
-
-# Set up minimal logging to see fallback messages
-logging.basicConfig(level=logging.DEBUG, format='%(levelname)s: %(message)s')
-
-try:
-    from praisonaiagents import Agent
-    
-    # Test 1: Create agent with Deepseek (a provider that typically doesn't support sync streaming)
-    print("🧪 Testing smart fallback with sync agent...")
-    
-    # Mock a scenario that would trigger the fallback
-    agent = Agent(
-        name="test_agent",
-        llm="deepseek-chat",  # This would typically trigger the fallback
-        instructions="You are a helpful assistant. Say hello."
-    )
-    
-    # Test _chat_completion with stream=None (should trigger smart fallback)
-    print("✨ Testing _chat_completion with stream=None (smart fallback)...")
-    
-    messages = [{"role": "user", "content": "Hello, please say hello back."}]
-    
-    # This should try streaming first, then fall back to non-streaming if needed
-    try:
-        response = agent._chat_completion(messages, stream=None)
-        print("✅ Smart fallback test passed - got response without errors")
-        if response and response.choices:
-            print(f"📝 Response: {response.choices[0].message.content[:100]}...")
-    except Exception as e:
-        print(f"❌ Smart fallback test failed: {e}")
-    
-    # Test 2: Explicit streaming should work if supported
-    print("\n✨ Testing _chat_completion with stream=True (explicit)...")
-    try:
-        response = agent._chat_completion(messages, stream=True)
-        print("✅ Explicit streaming test passed")
-    except ValueError as e:
-        if "Streaming is not supported" in str(e):
-            print("✅ Explicit streaming correctly rejected by adapter")
-        else:
-            print(f"❌ Unexpected error: {e}")
-    except Exception as e:
-        print(f"❌ Unexpected error: {e}")
-    
-    # Test 3: Explicit non-streaming should always work
-    print("\n✨ Testing _chat_completion with stream=False (explicit)...")
-    try:
-        response = agent._chat_completion(messages, stream=False)
-        print("✅ Explicit non-streaming test passed")
-    except Exception as e:
-        print(f"❌ Explicit non-streaming test failed: {e}")
-
-except ImportError as e:
-    print(f"❌ Import error: {e}")
-    print("Note: This is expected if dependencies are not installed")
-except Exception as e:
-    print(f"❌ Unexpected error: {e}")
-
-print("\n🎯 Smart fallback implementation test completed!")
\ No newline at end of file