agentic-blocks 0.1.8__tar.gz → 0.1.10__tar.gz

{agentic_blocks-0.1.8/src/agentic_blocks.egg-info → agentic_blocks-0.1.10}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: agentic-blocks
-Version: 0.1.8
+Version: 0.1.10
 Summary: Simple building blocks for agentic AI systems with MCP client and conversation management
 Author-email: Magnus Bjelkenhed <bjelkenhed@gmail.com>
 License: MIT
@@ -23,6 +23,7 @@ Requires-Dist: mcp
 Requires-Dist: requests
 Requires-Dist: python-dotenv
 Requires-Dist: openai
+Requires-Dist: langchain-core
 Provides-Extra: test
 Requires-Dist: pytest; extra == "test"
 Provides-Extra: dev

{agentic_blocks-0.1.8 → agentic_blocks-0.1.10}/pyproject.toml

@@ -14,7 +14,7 @@ agentic_blocks = []
 
 [project]
 name = "agentic-blocks"
-version = "0.1.8"
+version = "0.1.10"
 description = "Simple building blocks for agentic AI systems with MCP client and conversation management"
 readme = "README.md"
 requires-python = ">=3.11"
@@ -39,6 +39,7 @@ dependencies = [
     "requests",
     "python-dotenv",
     "openai",
+    "langchain-core",
 ]
 
 [project.urls]

{agentic_blocks-0.1.8 → agentic_blocks-0.1.10}/src/agentic_blocks/llm.py

@@ -9,6 +9,7 @@ from dotenv import load_dotenv
 from openai import OpenAI
 
 from agentic_blocks.messages import Messages
+from agentic_blocks.utils.tools_utils import langchain_tools_to_openai_format
 
 
 class LLMError(Exception):
@@ -19,7 +20,7 @@ class LLMError(Exception):
 
 def call_llm(
     messages: Union[Messages, List[Dict[str, Any]]],
-    tools: Optional[List[Dict[str, Any]]] = None,
+    tools: Optional[Union[List[Dict[str, Any]], List]] = None,
     api_key: Optional[str] = None,
     model: str = "gpt-4o-mini",
     base_url: Optional[str] = None,
@@ -30,7 +31,7 @@ def call_llm(
 
     Args:
         messages: Either a Messages instance or a list of message dicts
-        tools: Optional list of tools in OpenAI function calling format
+        tools: Optional list of tools in OpenAI function calling format or LangChain StructuredTools
         api_key: OpenAI API key (if not provided, loads from .env OPENAI_API_KEY)
         model: Model name to use for completion
         base_url: Base URL for the API (useful for VLLM or other OpenAI-compatible servers)
@@ -79,6 +80,15 @@ def call_llm(
     if not conversation_messages:
         raise LLMError("No messages provided for completion.")
 
+    # Handle tools parameter - convert LangChain tools if needed
+    openai_tools = None
+    if tools:
+        # Check if it's a list of LangChain StructuredTools
+        if tools and hasattr(tools[0], 'args_schema'):
+            openai_tools = langchain_tools_to_openai_format(tools)
+        else:
+            openai_tools = tools
+
     try:
         # Prepare completion parameters
         completion_params = {
@@ -87,8 +97,8 @@ def call_llm(
             **kwargs,
         }
 
-        if tools:
-            completion_params["tools"] = tools
+        if openai_tools:
+            completion_params["tools"] = openai_tools
             completion_params["tool_choice"] = "auto"
 
         # Make completion request

{agentic_blocks-0.1.8 → agentic_blocks-0.1.10}/src/agentic_blocks/messages.py

@@ -205,6 +205,51 @@ class Messages:
 
         return False
 
+    def get_pending_tool_calls(self) -> List[Dict[str, Any]]:
+        """
+        Get pending tool calls that need execution, formatted for MCPClient.call_tool().
+
+        Returns:
+            List of dictionaries with 'tool_name', 'arguments', and 'tool_call_id' keys
+        """
+        pending_calls = []
+        
+        if not self.messages:
+            return pending_calls
+
+        last_message = self.messages[-1]
+
+        # Check if the last message is an assistant message with tool calls
+        if last_message.get("role") == "assistant" and "tool_calls" in last_message:
+            # Get tool call IDs that have responses
+            responded_tool_call_ids = set()
+            for msg in reversed(self.messages):
+                if msg.get("role") == "tool" and msg.get("tool_call_id"):
+                    responded_tool_call_ids.add(msg.get("tool_call_id"))
+
+            # Find tool calls that don't have responses
+            for tool_call in last_message["tool_calls"]:
+                tool_call_id = tool_call.get("id")
+                if tool_call_id not in responded_tool_call_ids:
+                    function_info = tool_call.get("function", {})
+                    tool_name = function_info.get("name")
+                    arguments_str = function_info.get("arguments", "{}")
+                    
+                    # Parse arguments JSON string to dict
+                    import json
+                    try:
+                        arguments = json.loads(arguments_str)
+                    except json.JSONDecodeError:
+                        arguments = {}
+                    
+                    pending_calls.append({
+                        "tool_name": tool_name,
+                        "arguments": arguments,
+                        "tool_call_id": tool_call_id
+                    })
+
+        return pending_calls
+
     def __str__(self) -> str:
         """Return messages in a simple, readable format."""
         if not self.messages:

agentic_blocks-0.1.10/src/agentic_blocks/utils/tools_utils.py

@@ -0,0 +1,46 @@
+"""
+Utilities for working with tools across different formats.
+"""
+
+from typing import Dict, Any, List
+
+
+def langchain_tool_to_openai_format(tool) -> Dict[str, Any]:
+    """
+    Convert a LangChain StructuredTool to OpenAI function calling format.
+    
+    Args:
+        tool: A langchain_core.tools.structured.StructuredTool instance
+        
+    Returns:
+        Dictionary in OpenAI function calling format, compatible with 
+        MCPClient.list_tools() output and call_llm() tools parameter
+    """
+    schema = tool.args_schema.model_json_schema()
+    
+    return {
+        "type": "function",
+        "function": {
+            "name": schema.get("title", tool.name),
+            "description": schema.get("description", ""),
+            "parameters": {
+                "type": "object",
+                "properties": schema.get("properties", {}),
+                "required": schema.get("required", [])
+            }
+        }
+    }
+
+
+def langchain_tools_to_openai_format(tools: List) -> List[Dict[str, Any]]:
+    """
+    Convert a list of LangChain StructuredTools to OpenAI function calling format.
+    
+    Args:
+        tools: List of langchain_core.tools.structured.StructuredTool instances
+        
+    Returns:
+        List of dictionaries in OpenAI function calling format, compatible with 
+        MCPClient.list_tools() output and call_llm() tools parameter
+    """
+    return [langchain_tool_to_openai_format(tool) for tool in tools]

{agentic_blocks-0.1.8 → agentic_blocks-0.1.10/src/agentic_blocks.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: agentic-blocks
-Version: 0.1.8
+Version: 0.1.10
 Summary: Simple building blocks for agentic AI systems with MCP client and conversation management
 Author-email: Magnus Bjelkenhed <bjelkenhed@gmail.com>
 License: MIT
@@ -23,6 +23,7 @@ Requires-Dist: mcp
 Requires-Dist: requests
 Requires-Dist: python-dotenv
 Requires-Dist: openai
+Requires-Dist: langchain-core
 Provides-Extra: test
 Requires-Dist: pytest; extra == "test"
 Provides-Extra: dev

{agentic_blocks-0.1.8 → agentic_blocks-0.1.10}/src/agentic_blocks.egg-info/SOURCES.txt

@@ -9,4 +9,5 @@ src/agentic_blocks.egg-info/PKG-INFO
 src/agentic_blocks.egg-info/SOURCES.txt
 src/agentic_blocks.egg-info/dependency_links.txt
 src/agentic_blocks.egg-info/requires.txt
-src/agentic_blocks.egg-info/top_level.txt
+src/agentic_blocks.egg-info/top_level.txt
+src/agentic_blocks/utils/tools_utils.py

{agentic_blocks-0.1.8 → agentic_blocks-0.1.10}/src/agentic_blocks.egg-info/requires.txt

@@ -2,6 +2,7 @@ mcp
 requests
 python-dotenv
 openai
+langchain-core
 
 [dev]
 pytest