praisonaiagents 0.0.68__py3-none-any.whl → 0.0.69__py3-none-any.whl

praisonaiagents/llm/llm.py

@@ -436,15 +436,37 @@ class LLM:
                     
                     # Handle tool calls
                     if tool_calls and execute_tool_fn:
+                        # Convert tool_calls to a serializable format for all providers
+                        serializable_tool_calls = []
+                        for tc in tool_calls:
+                            if isinstance(tc, dict):
+                                serializable_tool_calls.append(tc)  # Already a dict
+                            else:
+                                # Convert object to dict
+                                serializable_tool_calls.append({
+                                    "id": tc.id,
+                                    "type": getattr(tc, 'type', "function"),
+                                    "function": {
+                                        "name": tc.function.name,
+                                        "arguments": tc.function.arguments
+                                    }
+                                })
                         messages.append({
                             "role": "assistant",
                             "content": response_text,
-                            "tool_calls": tool_calls
+                            "tool_calls": serializable_tool_calls
                         })
                         
                         for tool_call in tool_calls:
-                            function_name = tool_call["function"]["name"]
-                            arguments = json.loads(tool_call["function"]["arguments"])
+                            # Handle both object and dict access patterns
+                            if isinstance(tool_call, dict):
+                                function_name = tool_call["function"]["name"]
+                                arguments = json.loads(tool_call["function"]["arguments"])
+                                tool_call_id = tool_call["id"]
+                            else:
+                                function_name = tool_call.function.name
+                                arguments = json.loads(tool_call.function.arguments)
+                                tool_call_id = tool_call.id
 
                             logging.debug(f"[TOOL_EXEC_DEBUG] About to execute tool {function_name} with args: {arguments}")
                             tool_result = execute_tool_fn(function_name, arguments)
@@ -462,18 +484,11 @@ class LLM:
                                 logging.debug(f"[TOOL_EXEC_DEBUG] About to display tool call with message: {display_message}")
                                 display_tool_call(display_message, console=console)
                                 
-                                messages.append({
-                                    "role": "tool",
-                                    "tool_call_id": tool_call["id"],
-                                    "content": json.dumps(tool_result)
-                                })
-                            else:
-                                logging.debug("[TOOL_EXEC_DEBUG] Verbose mode off, not displaying tool call")
-                                messages.append({
-                                    "role": "tool",
-                                    "tool_call_id": tool_call["id"],
-                                    "content": "Function returned an empty output"
-                                })
+                            messages.append({
+                                "role": "tool",
+                                "tool_call_id": tool_call_id,
+                                "content": json.dumps(tool_result) if tool_result is not None else "Function returned an empty output"
+                            })
 
                         # If reasoning_steps is True, do a single non-streaming call
                         if reasoning_steps:
@@ -930,15 +945,37 @@ Output MUST be JSON with 'reflection' and 'satisfactory'.
                 tool_calls = tool_response.choices[0].message.get("tool_calls")
                 
                 if tool_calls:
+                    # Convert tool_calls to a serializable format for all providers
+                    serializable_tool_calls = []
+                    for tc in tool_calls:
+                        if isinstance(tc, dict):
+                            serializable_tool_calls.append(tc)  # Already a dict
+                        else:
+                            # Convert object to dict
+                            serializable_tool_calls.append({
+                                "id": tc.id,
+                                "type": getattr(tc, 'type', "function"),
+                                "function": {
+                                    "name": tc.function.name,
+                                    "arguments": tc.function.arguments
+                                }
+                            })
                     messages.append({
                         "role": "assistant",
                         "content": response_text,
-                        "tool_calls": tool_calls
+                        "tool_calls": serializable_tool_calls
                     })
                     
                     for tool_call in tool_calls:
-                        function_name = tool_call.function.name
-                        arguments = json.loads(tool_call.function.arguments)
+                        # Handle both object and dict access patterns
+                        if isinstance(tool_call, dict):
+                            function_name = tool_call["function"]["name"]
+                            arguments = json.loads(tool_call["function"]["arguments"])
+                            tool_call_id = tool_call["id"]
+                        else:
+                            function_name = tool_call.function.name
+                            arguments = json.loads(tool_call.function.arguments)
+                            tool_call_id = tool_call.id
 
                         tool_result = await execute_tool_fn(function_name, arguments)
 
@@ -949,17 +986,11 @@ Output MUST be JSON with 'reflection' and 'satisfactory'.
                             else:
                                 display_message += "Function returned no output"
                             display_tool_call(display_message, console=console)
-                            messages.append({
-                                "role": "tool",
-                                "tool_call_id": tool_call.id,
-                                "content": json.dumps(tool_result)
-                            })
-                        else:
-                            messages.append({
-                                "role": "tool",
-                                "tool_call_id": tool_call.id,
-                                "content": "Function returned an empty output"
-                            })
+                        messages.append({
+                            "role": "tool",
+                            "tool_call_id": tool_call_id,
+                            "content": json.dumps(tool_result) if tool_result is not None else "Function returned an empty output"
+                        })
 
                     # Get response after tool calls
                     response_text = ""

{praisonaiagents-0.0.68.dist-info → praisonaiagents-0.0.69.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: praisonaiagents
-Version: 0.0.68
+Version: 0.0.69
 Summary: Praison AI agents for completing complex tasks with Self Reflection Agents
 Author: Mervin Praison
 Requires-Dist: pydantic

{praisonaiagents-0.0.68.dist-info → praisonaiagents-0.0.69.dist-info}/RECORD

@@ -10,7 +10,7 @@ praisonaiagents/knowledge/__init__.py,sha256=xL1Eh-a3xsHyIcU4foOWF-JdWYIYBALJH9b
 praisonaiagents/knowledge/chunking.py,sha256=FzoNY0q8MkvG4gADqk4JcRhmH3lcEHbRdonDgitQa30,6624
 praisonaiagents/knowledge/knowledge.py,sha256=fQNREDiwdoisfIxJBLVkteXgq_8Gbypfc3UaZbxf5QY,13210
 praisonaiagents/llm/__init__.py,sha256=ttPQQJQq6Tah-0updoEXDZFKWtJAM93rBWRoIgxRWO8,689
-praisonaiagents/llm/llm.py,sha256=l7Z2QjD9eFy0Zq5bwTVK7VOUHxeTyx866YWt3fS3vz8,74606
+praisonaiagents/llm/llm.py,sha256=9UhKoTcInXfCFFmQfBEzs3G0t4dJ4yoEpQ5eUG9VC0Q,76574
 praisonaiagents/mcp/__init__.py,sha256=IkYdrAK1bDQDm_0t3Wjt63Zwv3_IJgqz84Wqz9GH2iQ,111
 praisonaiagents/mcp/mcp.py,sha256=BPPf5AIPXx28PaJJqOg6T3NRyymQH9YAD-Km7Ma9-KA,13681
 praisonaiagents/memory/memory.py,sha256=I8dOTkrl1i-GgQbDcrFOsSruzJ7MiI6Ys37DK27wrUs,35537
@@ -39,7 +39,7 @@ praisonaiagents/tools/xml_tools.py,sha256=iYTMBEk5l3L3ryQ1fkUnNVYK-Nnua2Kx2S0dxN
 praisonaiagents/tools/yaml_tools.py,sha256=uogAZrhXV9O7xvspAtcTfpKSQYL2nlOTvCQXN94-G9A,14215
 praisonaiagents/tools/yfinance_tools.py,sha256=s2PBj_1v7oQnOobo2fDbQBACEHl61ftG4beG6Z979ZE,8529
 praisonaiagents/tools/train/data/generatecot.py,sha256=H6bNh-E2hqL5MW6kX3hqZ05g9ETKN2-kudSjiuU_SD8,19403
-praisonaiagents-0.0.68.dist-info/METADATA,sha256=wMy9WDu6aGcQWrQIFT1UMOfB-wcvrGqpeDwXqbQXIvM,856
-praisonaiagents-0.0.68.dist-info/WHEEL,sha256=CmyFI0kx5cdEMTLiONQRbGQwjIoR1aIYB7eCAQ4KPJ0,91
-praisonaiagents-0.0.68.dist-info/top_level.txt,sha256=_HsRddrJ23iDx5TTqVUVvXG2HeHBL5voshncAMDGjtA,16
-praisonaiagents-0.0.68.dist-info/RECORD,,
+praisonaiagents-0.0.69.dist-info/METADATA,sha256=ymaq8EM2AZeNlbT-ISjYgHXMHbnHpX53O_Hkcrb5KtA,856
+praisonaiagents-0.0.69.dist-info/WHEEL,sha256=CmyFI0kx5cdEMTLiONQRbGQwjIoR1aIYB7eCAQ4KPJ0,91
+praisonaiagents-0.0.69.dist-info/top_level.txt,sha256=_HsRddrJ23iDx5TTqVUVvXG2HeHBL5voshncAMDGjtA,16
+praisonaiagents-0.0.69.dist-info/RECORD,,