alita-sdk 0.3.257__py3-none-any.whl → 0.3.584__py3-none-any.whl

alita_sdk/runtime/langchain/assistant.py

@@ -5,6 +5,7 @@ from typing import Any, Optional
 from langchain.agents import (
     AgentExecutor, create_openai_tools_agent,
     create_json_chat_agent)
+from langgraph.graph.state import CompiledStateGraph
 from langgraph.store.base import BaseStore
 from .agents.xml_chat import create_xml_chat_agent
 from .langraph_agent import create_graph
@@ -12,10 +13,14 @@ from langchain_core.messages import (
     BaseMessage, SystemMessage, HumanMessage
 )
 from langchain_core.prompts import MessagesPlaceholder
-from .constants import REACT_ADDON, REACT_VARS, XML_ADDON
+from .constants import (REACT_ADDON, REACT_VARS, XML_ADDON, USER_ADDON,
+                        QA_ASSISTANT, NERDY_ASSISTANT, QUIRKY_ASSISTANT, CYNICAL_ASSISTANT,
+                        DEFAULT_ASSISTANT, PLAN_ADDON, PYODITE_ADDON, DATA_ANALYSIS_ADDON,
+                        SEARCH_INDEX_ADDON, FILE_HANDLING_INSTRUCTIONS)
 from .chat_message_template import Jinja2TemplatedChatMessagesTemplate
 from ..tools.echo import EchoTool
 from langchain_core.tools import BaseTool, ToolException
+from jinja2 import Environment, DebugUndefined
 
 logger = logging.getLogger(__name__)
 
@@ -28,15 +33,23 @@ class Assistant:
                  app_type: str = "openai",
                  tools: Optional[list] = [],
                  memory: Optional[Any] = None,
-                 store: Optional[BaseStore] = None):
+                 store: Optional[BaseStore] = None,
+                 debug_mode: Optional[bool] = False,
+                 mcp_tokens: Optional[dict] = None,
+                 conversation_id: Optional[str] = None,
+                 ignored_mcp_servers: Optional[list] = None,
+                 persona: Optional[str] = "generic"):
 
         self.app_type = app_type
         self.memory = memory
         self.store = store
+        self.persona = persona
+        self.max_iterations = data.get('meta', {}).get('step_limit', 25)
 
         logger.debug("Data for agent creation: %s", data)
         logger.info("App type: %s", app_type)
-        
+
+        self.alita_client = alita
         self.client = client
         # For predict agents, use the client as-is since it's already configured
         # if app_type == "predict":
@@ -61,11 +74,11 @@ class Assistant:
         #     )
         #     self.client = target_cls(**model_params)
         # validate agents compatibility: non-pipeline agents cannot have pipelines as toolkits
-        if app_type not in ["pipeline", "predict"]:
-            tools_to_check = data.get('tools', [])
-            if any(tool['agent_type'] == 'pipeline' for tool in tools_to_check):
-                raise ToolException("Non-pipeline agents cannot have pipelines as a toolkits. "
-                                    "Review toolkits configuration or use pipeline as master agent.")
+        # if app_type not in ["pipeline", "predict"]:
+        #     tools_to_check = data.get('tools', [])
+        #     if any(tool['agent_type'] == 'pipeline' for tool in tools_to_check):
+        #         raise ToolException("Non-pipeline agents cannot have pipelines as a toolkits. "
+        #                             "Review toolkits configuration or use pipeline as master agent.")
 
         # configure memory store if memory tool is defined (not needed for predict agents)
         if app_type != "predict":
@@ -74,13 +87,88 @@ class Assistant:
         else:
             # For predict agents, initialize memory store to None since they don't use memory
             self.store = None
-        
+
         # Lazy import to avoid circular dependency
         from ..toolkits.tools import get_tools
-        
-        self.tools = get_tools(data['tools'], alita_client=alita, llm=self.client, memory_store=self.store)
+        version_tools = data['tools']
+        # Handle internal tools
+        meta = data.get('meta', {})
+        if meta.get("internal_tools"):
+            # Find bucket from artifact toolkit marked with is_attachment flag
+            bucket_name = None
+            for tool in version_tools:
+                if tool.get('type') == 'artifact' and tool.get('is_attachment'):
+                    bucket_name = tool.get('settings', {}).get('bucket')
+                    break
+            # Fallback: use first artifact toolkit with a bucket
+            if not bucket_name:
+                for tool in version_tools:
+                    if tool.get('type') == 'artifact' and tool.get('settings', {}).get('bucket'):
+                        bucket_name = tool['settings']['bucket']
+                        break
+
+            for internal_tool_name in meta.get("internal_tools"):
+                tool_config = {"type": "internal_tool", "name": internal_tool_name, "settings": {}}
+                if bucket_name:
+                    tool_config["settings"]["bucket_name"] = bucket_name
+                version_tools.append(tool_config)
+
+        self.tools = get_tools(
+            version_tools,
+            alita_client=alita,
+            llm=self.client,
+            memory_store=self.store,
+            debug_mode=debug_mode,
+            mcp_tokens=mcp_tokens,
+            conversation_id=conversation_id,
+            ignored_mcp_servers=ignored_mcp_servers
+        )
         if tools:
             self.tools += tools
+        
+        # Create ToolRegistry to track tool metadata and handle name collisions
+        self.tool_registry = {}
+        tool_name_counts = {}  # Track how many times each base name appears
+        
+        for tool in self.tools:
+            if hasattr(tool, 'name'):
+                original_name = tool.name
+                base_name = original_name
+                
+                # Extract toolkit metadata from tool configuration
+                toolkit_name = ""
+                toolkit_type = ""
+                
+                # Find matching tool config to extract metadata
+                for tool_config in version_tools:
+                    # Try to match by toolkit_name or name field
+                    config_toolkit_name = tool_config.get('toolkit_name', tool_config.get('name', ''))
+                    # Simple heuristic: toolkit info should be accessible from tool config
+                    # For now, use toolkit_name and type from config
+                    toolkit_name = config_toolkit_name
+                    toolkit_type = tool_config.get('type', '')
+                    break  # Use first match for now; will refine with better matching
+                
+                # Handle duplicate tool names by appending numeric suffix
+                if base_name in tool_name_counts:
+                    tool_name_counts[base_name] += 1
+                    # Append suffix to make unique
+                    new_name = f"{base_name}_{tool_name_counts[base_name]}"
+                    tool.name = new_name
+                    logger.info(f"Tool name collision detected: '{base_name}' -> '{new_name}'")
+                else:
+                    tool_name_counts[base_name] = 0
+                    new_name = base_name
+                
+                # Store in registry
+                self.tool_registry[tool.name] = {
+                    'toolkit_name': toolkit_name,
+                    'toolkit_type': toolkit_type,
+                    'original_tool_name': base_name
+                }
+                
+        logger.info(f"ToolRegistry initialized with {len(self.tool_registry)} tools")
+        
         # Handle prompt setup
         if app_type in ["pipeline", "predict", "react"]:
             self.prompt = data['instructions']
@@ -92,7 +180,7 @@ class Assistant:
             elif app_type == "xml":
                 messages.append(HumanMessage(XML_ADDON))
             elif app_type in ['openai', 'dial']:
-                messages.append(HumanMessage("{{input}}"))
+                messages.append(MessagesPlaceholder("input"))
             messages.append(MessagesPlaceholder("agent_scratchpad"))
             variables = {}
             input_variables = []
@@ -108,13 +196,18 @@ class Assistant:
                 messages.extend(chat_history)
             self.prompt = Jinja2TemplatedChatMessagesTemplate(messages=messages)
             if input_variables:
-                self.prompt.input_variables = input_variables
+                if hasattr(self.prompt, 'input_variables') and self.prompt.input_variables is not None:
+                    self.prompt.input_variables.extend(input_variables)
+                else:
+                    self.prompt.input_variables = input_variables
             if variables:
                 self.prompt.partial_variables = variables
             try:
-                logger.info(f"Client was created with client setting: temperature - {self.client._get_model_default_parameters}")
+                logger.info(
+                    f"Client was created with client setting: temperature - {self.client._get_model_default_parameters}")
             except Exception as e:
-                logger.info(f"Client was created with client setting: temperature - {self.client.temperature} : {self.client.max_tokens}")
+                logger.info(
+                    f"Client was created with client setting: temperature - {self.client.temperature} : {self.client.max_tokens}")
 
     def _configure_store(self, memory_tool: dict | None) -> None:
         """
@@ -124,18 +217,16 @@ class Assistant:
         if not memory_tool or self.store is not None:
             return
         from .store_manager import get_manager
-        conn_str = memory_tool.get('settings', {}).get('connection_string', '')
+        conn_str = memory_tool['settings'].get('pgvector_configuration', {}).get('connection_string', '')
         store = get_manager().get_store(conn_str)
         self.store = store
 
     def runnable(self):
         if self.app_type == 'pipeline':
             return self.pipeline()
-        elif self.app_type == 'openai':
-            return self.getOpenAIToolsAgentExecutor()
         elif self.app_type == 'xml':
             return self.getXMLAgentExecutor()
-        elif self.app_type in ['predict', 'react']:
+        elif self.app_type in ['predict', 'react', 'openai']:
             return self.getLangGraphReactAgent()
         else:
             self.tools = [EchoTool()] + self.tools
@@ -144,24 +235,24 @@ class Assistant:
     def _agent_executor(self, agent: Any):
         return AgentExecutor.from_agent_and_tools(agent=agent, tools=self.tools,
                                                   verbose=True, handle_parsing_errors=True,
-                                                  max_execution_time=None, return_intermediate_steps=True)
+                                                  max_execution_time=None, return_intermediate_steps=True,
+                                                  max_iterations=self.max_iterations)
 
     def getAgentExecutor(self):
         # Exclude compiled graph runnables from simple tool agents
-        simple_tools = [t for t in self.tools if isinstance(t, BaseTool)]
+        simple_tools = [t for t in self.tools if isinstance(t, (BaseTool, CompiledStateGraph))]
         agent = create_json_chat_agent(llm=self.client, tools=simple_tools, prompt=self.prompt)
         return self._agent_executor(agent)
 
-
     def getXMLAgentExecutor(self):
         # Exclude compiled graph runnables from simple tool agents
-        simple_tools = [t for t in self.tools if isinstance(t, BaseTool)]
+        simple_tools = [t for t in self.tools if isinstance(t, (BaseTool, CompiledStateGraph))]
         agent = create_xml_chat_agent(llm=self.client, tools=simple_tools, prompt=self.prompt)
         return self._agent_executor(agent)
 
     def getOpenAIToolsAgentExecutor(self):
         # Exclude compiled graph runnables from simple tool agents
-        simple_tools = [t for t in self.tools if isinstance(t, BaseTool)]
+        simple_tools = [t for t in self.tools if isinstance(t, (BaseTool, CompiledStateGraph))]
         agent = create_openai_tools_agent(llm=self.client, tools=simple_tools, prompt=self.prompt)
         return self._agent_executor(agent)
 
@@ -171,7 +262,7 @@ class Assistant:
         This creates a proper LangGraphAgentRunnable with modern tool support.
         """
         # Exclude compiled graph runnables from simple tool agents
-        simple_tools = [t for t in self.tools if isinstance(t, BaseTool)]
+        simple_tools = [t for t in self.tools if isinstance(t, (BaseTool, CompiledStateGraph))]
         
         # Set up memory/checkpointer if available
         checkpointer = None
@@ -207,29 +298,43 @@ class Assistant:
         if self.app_type == "predict" and isinstance(self.prompt, str):
             prompt_instructions = self.prompt
         
-        # Create a unified YAML schema with conditional tool binding
-        # Build the base node configuration
-        node_config = {
-            'id': 'agent',
-            'type': 'llm',
-            'prompt': {
-                'template': prompt_instructions or "You are a helpful assistant."
-            },
-            'input': ['messages'],
-            'output': ['messages'],
-            'transition': 'END'
-        }
-        
         # Add tool binding only if tools are present
+        tool_names = []
         if simple_tools:
             tool_names = [tool.name for tool in simple_tools]
-            tool_names_yaml = str(tool_names).replace("'", '"')  # Convert to YAML-compatible format
-            node_config['tool_names'] = tool_names_yaml
             logger.info("Binding tools: %s", tool_names)
         
+        # take the system message from the openai prompt as a prompt instructions
+        if self.app_type == "openai" and hasattr(self.prompt, 'messages'):
+            prompt_instructions = self.__take_prompt_from_openai_messages()
+        
+        user_addon = USER_ADDON.format(prompt=str(prompt_instructions)) if prompt_instructions else ""
+        plan_addon = PLAN_ADDON if 'update_plan' in tool_names else ""
+        data_analysis_addon = DATA_ANALYSIS_ADDON if 'pandas_analyze_data' in tool_names else ""
+        pyodite_addon = PYODITE_ADDON if 'pyodide_sandbox' in tool_names else ""
+        search_index_addon = SEARCH_INDEX_ADDON if 'stepback_summary_index' in tool_names else ""
+
+        # Select assistant template based on persona
+        persona_templates = {
+            "qa": QA_ASSISTANT,
+            "nerdy": NERDY_ASSISTANT,
+            "quirky": QUIRKY_ASSISTANT,
+            "cynical": CYNICAL_ASSISTANT,
+        }
+        base_assistant = persona_templates.get(self.persona, DEFAULT_ASSISTANT)
+
+        escaped_prompt = base_assistant.format(
+            users_instructions=user_addon,
+            planning_instructions=plan_addon,
+            pyodite_addon=pyodite_addon,
+            data_analysis_addon=data_analysis_addon,
+            search_index_addon=search_index_addon,
+            file_handling_instructions=FILE_HANDLING_INSTRUCTIONS
+        )
+
         # Properly setup the prompt for YAML
         import yaml
-        escaped_prompt = prompt_instructions or "You are a helpful assistant."
+        
         
         # Create the schema as a dictionary first, then convert to YAML
         state_messages_config = {'type': 'list'}
@@ -247,6 +352,9 @@ class Assistant:
         schema_dict = {
             'name': 'react_agent',
             'state': {
+                'input': {
+                    'type': 'str'
+                },
                 'messages': state_messages_config
             },
             'nodes': [{
@@ -255,6 +363,21 @@ class Assistant:
                 'prompt': {
                     'template': escaped_prompt
                 },
+                'input_mapping': {
+                    'system': {
+                        'type': 'fixed',
+                        'value': escaped_prompt
+                    },
+                    'task': {
+                        'type': 'variable',
+                        'value': 'input'
+                    },
+                    'chat_history': {
+                        'type': 'variable',
+                        'value': 'messages'
+                    }
+                },
+                'step_limit': self.max_iterations,
                 'input': ['messages'],
                 'output': ['messages'],
                 'transition': 'END'
@@ -279,7 +402,9 @@ class Assistant:
             memory=checkpointer,
             store=self.store,
             debug=False,
-            for_subgraph=False
+            for_subgraph=False,
+            alita_client=self.alita_client,
+            steps_limit=self.max_iterations
         )
         
         return agent
@@ -293,7 +418,9 @@ class Assistant:
         #
         agent = create_graph(
             client=self.client, tools=self.tools,
-            yaml_schema=self.prompt, memory=memory
+            yaml_schema=self.prompt, memory=memory,
+            alita_client=self.alita_client,
+            steps_limit=self.max_iterations
         )
         #
         return agent
@@ -304,3 +431,16 @@ class Assistant:
 
     def predict(self, messages: list[BaseMessage]):
         return self.client.invoke(messages)
+
+    def __take_prompt_from_openai_messages(self):
+        if self.prompt and self.prompt.messages:
+            for message in self.prompt.messages:
+                # we don't need any message placeholder from the openai agent prompt
+                if hasattr(message, 'variable_name'):
+                    continue
+                # take only the content of the system message from the openai prompt
+                if isinstance(message, SystemMessage):
+                    environment = Environment(undefined=DebugUndefined)
+                    template = environment.from_string(message.content)
+                    return template.render(self.prompt.partial_variables)
+        return None