agentUniverse 0.0.17.dev1__tar.gz → 0.0.17.dev2__tar.gz

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/PKG-INFO

@@ -1,8 +1,7 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.3
 Name: agentUniverse
-Version: 0.0.17.dev1
+Version: 0.0.17.dev2
 Summary: agentUniverse is a framework for developing applications powered by multi-agent base on large language model.
-Home-page: https://github.com/agentuniverse-ai/agentUniverse
 Author: AntGroup
 Author-email: jerry.zzw@antgroup.com
 Requires-Python: >=3.10,<4.0
@@ -12,6 +11,7 @@ Classifier: Programming Language :: Python :: 3
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
 Provides-Extra: log-ext
 Provides-Extra: store-ext
 Requires-Dist: Jinja2 (>=3.1.4,<4.0.0)
@@ -45,16 +45,15 @@ Requires-Dist: pandas (>=2.2.2,<3.0.0)
 Requires-Dist: pillow (>=10.4.0,<11.0.0)
 Requires-Dist: primp (>=0.6.5,<0.7.0)
 Requires-Dist: pyarrow (>=16.1.0,<17.0.0)
-Requires-Dist: pydantic (>=2.10.0,<3.0.0)
+Requires-Dist: pydantic (>=2.6.4,<3.0.0)
 Requires-Dist: pymilvus (>=2.4.3,<3.0.0) ; extra == "store-ext"
 Requires-Dist: qianfan (>=0.3.12,<0.4.0)
 Requires-Dist: requests (>=2.32.0,<3.0.0)
 Requires-Dist: sphinx (>=7.2.6,<8.0.0)
 Requires-Dist: sphinx-rtd-theme (>=2.0.0,<3.0.0)
-Requires-Dist: tiktoken (==0.5.2)
+Requires-Dist: tiktoken (<1.0.0)
 Requires-Dist: tomli (>=2.2,<3.0)
 Requires-Dist: tqdm (>=4.66.3,<5.0.0)
-Requires-Dist: tree-sitter (>=0.24.0,<0.25.0)
 Requires-Dist: werkzeug (>=3.0.3,<4.0.0)
 Requires-Dist: wikipedia (>=1.4.0,<2.0.0)
 Project-URL: Repository, https://github.com/agentuniverse-ai/agentUniverse
@@ -63,9 +62,9 @@ Description-Content-Type: text/markdown
 # agentUniverse
 
 ## Overview
-agentUniverse is a framework for developing applications powered by multi-agent base on large language model.  It provides all the essential components for building a single agent, and a multi-agent collaboration mechanism which  serves as a pattern factory that allowing developers to buid and customize multi-agent collaboration patterns. With this framework,  developers can easily construct multi-agent applications, and share the pattern practices from different technical  and business fields.
+agentUniverse is a framework for developing applications powered by multi-agent base on large language model.  It provides all the essential components for building a single agent, and a multi-agent collaboration mechanism which  serves as a pattern factory that allowing developers to build and customize multi-agent collaboration patterns. With this framework,  developers can easily construct multi-agent applications, and share the pattern practices from different technical  and business fields.
 
-The framework will come with serveral pre-install multi-agent collaboration patterns which have been proven effective in real business scenarios, and will continue to be enriched in the future. Patterns that are currently about to be released include:
+The framework will come with several pre-install multi-agent collaboration patterns which have been proven effective in real business scenarios, and will continue to be enriched in the future. Patterns that are currently about to be released include:
 
 - PEER pattern：
 This pattern utilizes four distinct agent roles: Plan, Execute, Express, and Review, to achieve a multi-step breakdown and sequential execution of a complex task. It also performs autonomous iteration based on evaluative feedback which enhancing performance in reasoning and analytical tasks. 

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/README_PYPI.md

@@ -1,9 +1,9 @@
 # agentUniverse
 
 ## Overview
-agentUniverse is a framework for developing applications powered by multi-agent base on large language model.  It provides all the essential components for building a single agent, and a multi-agent collaboration mechanism which  serves as a pattern factory that allowing developers to buid and customize multi-agent collaboration patterns. With this framework,  developers can easily construct multi-agent applications, and share the pattern practices from different technical  and business fields.
+agentUniverse is a framework for developing applications powered by multi-agent base on large language model.  It provides all the essential components for building a single agent, and a multi-agent collaboration mechanism which  serves as a pattern factory that allowing developers to build and customize multi-agent collaboration patterns. With this framework,  developers can easily construct multi-agent applications, and share the pattern practices from different technical  and business fields.
 
-The framework will come with serveral pre-install multi-agent collaboration patterns which have been proven effective in real business scenarios, and will continue to be enriched in the future. Patterns that are currently about to be released include:
+The framework will come with several pre-install multi-agent collaboration patterns which have been proven effective in real business scenarios, and will continue to be enriched in the future. Patterns that are currently about to be released include:
 
 - PEER pattern：
 This pattern utilizes four distinct agent roles: Plan, Execute, Express, and Review, to achieve a multi-step breakdown and sequential execution of a complex task. It also performs autonomous iteration based on evaluative feedback which enhancing performance in reasoning and analytical tasks. 

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/agentuniverse/agent/action/knowledge/doc_processor/character_text_splitter.py

@@ -31,13 +31,17 @@ class CharacterTextSplitter(DocProcessor):
     chunk_size: int = 200
     chunk_overlap: int = 20
     separator: str = "/n/n"
-    splitter: Optional[Splitter] = None
+    __splitter: Optional[Splitter] = None
 
-    def __init__(self, **kwargs):
-        super().__init__(**kwargs)
-        self.splitter = Splitter(separator=self.separator,
+
+    @property
+    def splitter(self) -> Splitter:
+        if not self.__splitter:
+            self.__splitter = Splitter(separator=self.separator,
                                  chunk_size=self.chunk_size,
                                  chunk_overlap=self.chunk_overlap)
+        return self.__splitter
+
 
     def _process_docs(self, origin_docs: List[Document], query: Query = None) -> \
             List[Document]:
@@ -72,7 +76,4 @@ class CharacterTextSplitter(DocProcessor):
             self.chunk_overlap = doc_processor_configer.chunk_overlap
         if hasattr(doc_processor_configer, "separator"):
             self.separator = doc_processor_configer.separator
-        self.splitter = Splitter(separator=self.separator,
-                                 chunk_size=self.chunk_size,
-                                 chunk_overlap=self.chunk_overlap)
         return self

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/agentuniverse/agent/action/knowledge/doc_processor/recursive_character_text_splitter.py

@@ -22,13 +22,15 @@ class RecursiveCharacterTextSplitter(DocProcessor):
     chunk_size: int = 200
     chunk_overlap: int = 20
     separators: List[str] = ["\n\n", "\n", " ", ""]
-    splitter: Optional[Splitter] = None
+    __splitter: Optional[Splitter] = None
 
-    def __init__(self, **kwargs):
-        super().__init__(**kwargs)
-        self.splitter = Splitter(separators=self.separators,
-                                 chunk_size=self.chunk_size,
-                                 chunk_overlap=self.chunk_overlap)
+    @property
+    def splitter(self) -> Splitter:
+        if not self.__splitter:
+            self.__splitter = Splitter(separator=self.separator,
+                                       chunk_size=self.chunk_size,
+                                       chunk_overlap=self.chunk_overlap)
+        return self.__splitter
 
     def _process_docs(self, origin_docs: List[Document], query: Query = None) -> \
             List[Document]:
@@ -63,9 +65,6 @@ class RecursiveCharacterTextSplitter(DocProcessor):
             self.chunk_overlap = doc_processor_configer.chunk_overlap
         if hasattr(doc_processor_configer, "separators"):
             self.separators = doc_processor_configer.separators
-        self.splitter = Splitter(separators=self.separators,
-                                 chunk_size=self.chunk_size,
-                                 chunk_overlap=self.chunk_overlap)
         return self
 
 

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/agentuniverse/agent/action/knowledge/doc_processor/token_text_splitter.py

@@ -22,16 +22,15 @@ class TokenTextSplitter(DocProcessor):
     chunk_overlap: int = 20
     encoding_name: str = 'gpt2'
     model_name: Optional[str] = None
-    splitter: Optional[Splitter] = None
+    __splitter: Optional[Splitter] = None
 
-    def __init__(self, **kwargs):
-        super().__init__(**kwargs)
-        self.splitter = Splitter(
-            encoding_name=self.encoding_name,
-            model_name=self.model_name,
-            chunk_size=self.chunk_size,
-            chunk_overlap=self.chunk_overlap
-        )
+    @property
+    def splitter(self) -> Splitter:
+        if not self.__splitter:
+            self.__splitter = Splitter(separator=self.separator,
+                                       chunk_size=self.chunk_size,
+                                       chunk_overlap=self.chunk_overlap)
+        return self.__splitter
 
     def _process_docs(self, origin_docs: List[Document], query: Query = None) -> \
             List[Document]:
@@ -68,10 +67,4 @@ class TokenTextSplitter(DocProcessor):
             self.encoding_name = doc_processor_configer.encoding_name
         if hasattr(doc_processor_configer, "model_name"):
             self.model_name = doc_processor_configer.model_name
-        self.splitter = Splitter(
-            encoding_name=self.encoding_name,
-            model_name=self.model_name,
-            chunk_size=self.chunk_size,
-            chunk_overlap=self.chunk_overlap
-        )
         return self

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/agentuniverse/agent/action/knowledge/knowledge.py

@@ -1,6 +1,6 @@
 # !/usr/bin/env python3
 # -*- coding:utf-8 -*-
-
+import asyncio
 # @Time    : 2024/3/22 15:44
 # @Author  : wangchongshi
 # @Email   : wangchongshi.wcs@antgroup.com
@@ -94,14 +94,14 @@ class Knowledge(ComponentBase):
             thread_name_prefix="Knowledge query"
         )
 
-    def _load_data(self,  *args: Any, **kwargs: Any) -> List[Document]:
+    def _load_data(self, *args: Any, **kwargs: Any) -> List[Document]:
         # check if source is a local file or remote url
         if kwargs.get("source_path"):
             source_path = kwargs.get("source_path")
         else:
             raise Exception("No file to load.")
         url_pattern = re.compile(
-            r'^(https?:\/\/)?' 
+            r'^(https?:\/\/)?'
             r'((([a-zA-Z0-9]{1,256}\.[a-zA-Z0-9]{1,6})|'
             r'(\d{1,3}\.){3}\d{1,3})'
             r'(:\d{1,5})?)'
@@ -277,7 +277,16 @@ class Knowledge(ComponentBase):
         """
         parse_query = parse_json_markdown(query)
         knowledge = self.query_knowledge(**parse_query)
-        return "This is Query Result:\n"+self.to_llm(knowledge)
+        return "This is Query Result:\n" + self.to_llm(knowledge)
+
+    async def async_langchain_query(self, query: str) -> str:
+        """Query the knowledge using LangChain.
+
+        Query documents from the store and return the results.
+        """
+        parse_query = parse_json_markdown(query)
+        knowledge = await asyncio.to_thread(self.query_knowledge, **parse_query)
+        return "This is Query Result:\n" + self.to_llm(knowledge)
 
     def as_langchain_tool(self) -> LangchainTool:
         """Convert the Knowledge object to a LangChain tool.
@@ -300,6 +309,27 @@ class Knowledge(ComponentBase):
             func=self.langchain_query,
         )
 
+    async def async_as_langchain_tool(self) -> LangchainTool:
+        """Convert the Knowledge object to a LangChain tool.
+
+        Returns:
+            Any: the LangChain tool object
+        """
+        args_description = """
+        This is a knowledge base tool, which stores the content you may need. To use this tool, you need to give a json string with the following format:
+        ```json
+        {
+            "query_str": "<your query here>",
+            "similarity_top_k": <number of results to return>,
+        }
+        ```
+        """
+        return LangchainTool(
+            name=self.name,
+            description=self.description or '' + args_description,
+            func=self.async_langchain_query,
+        )
+
     def create_copy(self):
         copied = self.model_copy()
         copied.stores = self.stores.copy()

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/agentuniverse/agent/action/tool/common_tool/google_search_tool.py

@@ -29,3 +29,9 @@ class GoogleSearchTool(Tool):
         # get top10 results from Google search.
         search = GoogleSerperAPIWrapper(serper_api_key=self.serper_api_key, k=10, gl="us", hl="en", type="search")
         return search.run(query=input)
+
+    async def async_execute(self, tool_input: ToolInput):
+        input = tool_input.get_data("input")
+        # get top10 results from Google search.
+        search = GoogleSerperAPIWrapper(serper_api_key=self.serper_api_key, k=10, gl="us", hl="en", type="search")
+        return await search.arun(query=input)

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/agentuniverse/agent/action/tool/common_tool/langchain_tool.py

@@ -23,6 +23,11 @@ class LangChainTool(Tool):
         callbacks = tool_input.get_data("callbacks", None)
         return self.tool.run(input, callbacks=callbacks)
 
+    async def async_execute(self, **kwargs):
+        input = kwargs.get("input")
+        callbacks = kwargs.get("callbacks", None)
+        return await self.tool.arun(input, callbacks=callbacks)
+
     def initialize_by_component_configer(self, component_configer: ToolConfiger) -> 'Tool':
         super().initialize_by_component_configer(component_configer)
         self.tool = self.init_langchain_tool(component_configer)

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/agentuniverse/agent/action/tool/common_tool/mock_search_tool.py

@@ -19,6 +19,12 @@ class MockSearchTool(Tool):
         res = self.mock_api_res()
         return res
 
+    async def async_execute(self, tool_input: ToolInput):
+        input = tool_input.get_data("input")
+        # get top10 results from mock search.
+        res = self.mock_api_res()
+        return res
+
     def mock_api_res(self):
         res = f"""采访中谈及第十次减持比亚迪，巴菲特称是为了便于伯克希尔进行更好的资金配置。而在今年2月，
         芒格在美国报纸和软件公司Daily Journal举行的虚拟年会上也曾谈及减持比亚迪的 ... 

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/agentuniverse/agent/action/tool/common_tool/request_tool.py

@@ -40,6 +40,19 @@ class RequestTool(Tool):
         else:
             return self.execute_by_method(input_params)
 
+    async def async_execute_by_method(self, url: str, data: dict = None, **kwargs):
+        url = self._clean_url(url)
+        if self.method == 'GET':
+            return await self.requests_wrapper.aget(url)
+        elif self.method == 'POST':
+            return await self.requests_wrapper.apost(url, data=data)
+        elif self.method == 'PUT':
+            return await self.requests_wrapper.aput(url, data=data)
+        elif self.method == 'DELETE':
+            return await self.requests_wrapper.adelete(url)
+        else:
+            raise ValueError(f"Unsupported method: {self.method}")
+
     def execute_by_method(self, url: str, data: dict = None, **kwargs):
         url = self._clean_url(url)
         if self.method == 'GET':

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/agentuniverse/agent/action/tool/common_tool/search_api_tool.py

@@ -57,6 +57,19 @@ class SearchAPITool(Tool):
             return self.search_api_wrapper.results(query=input, **search_params)
         return self.search_api_wrapper.run(query=input, **search_params)
 
+    async def async_execute(self, tool_input: ToolInput):
+        self._load_api_wapper()
+        search_params = {}
+        for k, v in self.search_params.items():
+            if k in tool_input.to_dict():
+                search_params[k] = tool_input.get_data(k)
+                continue
+            search_params[k] = v
+        input = tool_input.get_data("input")
+        if self.search_type == "json":
+            return await self.search_api_wrapper.aresults(query=input, **search_params)
+        return await self.search_api_wrapper.arun(query=input, **search_params)
+
     def initialize_by_component_configer(self, component_configer: ToolConfiger) -> 'Tool':
         """Initialize the tool by the component configer."""
         super().initialize_by_component_configer(component_configer)

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/agentuniverse/agent/action/tool/common_tool/simple_math_tool.py

@@ -17,6 +17,12 @@ class AddTool(Tool):
         result = float(a) + float(b)
         return result
 
+    async def async_execute(self, tool_input: ToolInput):
+        input_params = tool_input.get_data('input')
+        a, b = input_params.split(',')
+        result = float(a) + float(b)
+        return result
+
 
 class SubtractTool(Tool):
     def execute(self, tool_input: ToolInput):
@@ -25,6 +31,12 @@ class SubtractTool(Tool):
         result = float(a) - float(b)
         return result
 
+    async def async_execute(self, tool_input: ToolInput):
+        input_params = tool_input.get_data('input')
+        a, b = input_params.split(',')
+        result = float(a) - float(b)
+        return result
+
 
 class MultiplyTool(Tool):
     def execute(self, tool_input: ToolInput):
@@ -33,6 +45,12 @@ class MultiplyTool(Tool):
         result = float(a) * float(b)
         return result
 
+    async def async_execute(self, tool_input: ToolInput):
+        input_params = tool_input.get_data('input')
+        a, b = input_params.split(',')
+        result = float(a) * float(b)
+        return result
+
 
 class DivideTool(Tool):
     def execute(self, tool_input: ToolInput):
@@ -40,3 +58,9 @@ class DivideTool(Tool):
         a, b = input_params.split(',')
         result = float(a) / float(b)
         return result
+
+    async def async_execute(self, tool_input: ToolInput):
+        input_params = tool_input.get_data('input')
+        a, b = input_params.split(',')
+        result = float(a) / float(b)
+        return result

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/agentuniverse/agent/action/tool/tool.py

@@ -1,6 +1,6 @@
 # !/usr/bin/env python3
 # -*- coding:utf-8 -*-
-
+import asyncio
 # @Time    : 2024/3/13 14:29
 # @Author  : wangchongshi
 # @Email   : wangchongshi.wcs@antgroup.com
@@ -70,6 +70,13 @@ class Tool(ComponentBase):
         tool_input = ToolInput(kwargs)
         return self.execute(tool_input)
 
+    @trace_tool
+    async def async_run(self, **kwargs):
+        """The callable method that runs the tool."""
+        self.input_check(kwargs)
+        tool_input = ToolInput(kwargs)
+        return await self.async_execute(tool_input)
+
     def input_check(self, kwargs: dict) -> None:
         """Check whether the input parameters of the tool contain input keys of the tool"""
         if self.input_keys:
@@ -87,6 +94,15 @@ class Tool(ComponentBase):
             tool_input.add_data(key, parse_result[key])
         return self.execute(tool_input)
 
+    @trace_tool
+    async def async_langchain_run(self, *args, callbacks=None, **kwargs):
+        kwargs["callbacks"] = callbacks
+        tool_input = ToolInput(kwargs)
+        parse_result = self.parse_react_input(args[0])
+        for key in self.input_keys:
+            tool_input.add_data(key, parse_result[key])
+        return await self.async_execute(tool_input)
+
     def parse_react_input(self, input_str: str):
         """
             parse react string to you input
@@ -100,12 +116,22 @@ class Tool(ComponentBase):
     def execute(self, tool_input: ToolInput):
         raise NotImplementedError
 
+    async def async_execute(self, tool_input: ToolInput):
+        """The callable method that runs the tool."""
+        return await asyncio.to_thread(self.execute, tool_input)
+
     def as_langchain(self) -> LangchainTool:
         """Convert the agentUniverse(aU) tool class to the langchain tool class."""
         return LangchainTool(name=self.name,
                              func=self.langchain_run,
                              description=self.description)
 
+    async def async_as_langchain(self) -> LangchainTool:
+        return LangchainTool(name=self.name,
+                             func=self.run,
+                             coroutine=self.async_langchain_run,
+                             description=self.description)
+
     def get_instance_code(self) -> str:
         """Return the full name of the tool."""
         appname = ApplicationConfigManager().app_configer.base_info_appname

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/agentuniverse/agent/action/tool/tool_manager.py

@@ -39,8 +39,14 @@ class ToolManager(ComponentManagerBase):
                 tool_configer = tool_configer_map.get(component_instance_name)
                 if tool_configer:
                     # Dynamically import the module and retrieve the class specified in the configuration
-                    module = importlib.import_module(tool_configer.metadata_module)
-                    component_clz = getattr(module, tool_configer.metadata_class)
+                    if tool_configer.meta_class:
+                        metadata_module = '.'.join(tool_configer.meta_class.split('.')[:-1])
+                        metadata_class = tool_configer.meta_class.split('.')[-1]
+                        module = importlib.import_module(metadata_module)
+                        component_clz = getattr(module, metadata_class)
+                    else:
+                        module = importlib.import_module(tool_configer.metadata_module)
+                        component_clz = getattr(module, tool_configer.metadata_class)
                     # Initialize the tool instance using the configuration
                     instance_obj: Tool = component_clz().initialize_by_component_configer(tool_configer)
                     if instance_obj:

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/agentuniverse/agent/agent.py

@@ -42,6 +42,7 @@ from agentuniverse.base.context.framework_context_manager import FrameworkContex
 from agentuniverse.base.util.logging.logging_util import LOGGER
 from agentuniverse.base.util.memory_util import generate_messages, get_memory_string
 from agentuniverse.base.util.system_util import process_dict_with_funcs, is_system_builtin
+from agentuniverse.base.util.tracing.au_trace_manager import AuTraceManager
 from agentuniverse.llm.llm import LLM
 from agentuniverse.llm.llm_manager import LLMManager
 from agentuniverse.prompt.chat_prompt import ChatPrompt
@@ -54,6 +55,7 @@ class Agent(ComponentBase, ABC):
     """The parent class of all agent models, containing only attributes."""
 
     agent_model: Optional[AgentModel] = None
+    scene_code: Optional[str] = None
 
     def __init__(self):
         """Initialize the AgentModel with the given keyword arguments."""
@@ -92,6 +94,15 @@ class Agent(ComponentBase, ABC):
         """
         pass
 
+    def update_trace_context(self, input_object: InputObject):
+        session_id = input_object.get_data("session_id")
+        if session_id:
+            AuTraceManager().trace_context.set_session_id(session_id)
+
+        trace_id = input_object.get_data("trace_id")
+        if trace_id:
+            AuTraceManager().trace_context.set_trace_id(trace_id)
+
     @trace_agent
     def run(self, **kwargs) -> OutputObject:
         """Agent instance running entry.
@@ -102,6 +113,8 @@ class Agent(ComponentBase, ABC):
         self.input_check(kwargs)
         input_object = InputObject(kwargs)
 
+        self.update_trace_context(input_object)
+
         agent_input = self.pre_parse_input(input_object)
 
         planner_result = self.execute(input_object, agent_input)
@@ -116,6 +129,7 @@ class Agent(ComponentBase, ABC):
     async def async_run(self, **kwargs) -> OutputObject:
         self.input_check(kwargs)
         input_object = InputObject(kwargs)
+        self.update_trace_context(input_object)
 
         agent_input = self.pre_parse_input(input_object)
 
@@ -220,6 +234,19 @@ class Agent(ComponentBase, ABC):
             result_dict[key] = output_object.get_data(key)
         return result_dict
 
+    async def async_langchain_run(self, input: str, callbacks=None, **kwargs):
+        """Run the agent model using LangChain."""
+        try:
+            parse_result = parse_json_markdown(input)
+        except Exception as e:
+            LOGGER.error(f"langchain run parse_json_markdown error,input(parse_result) error({str(e)})")
+            return "Error , Your Action Input is not a valid JSON string"
+        output_object = await self.async_run(**parse_result, callbacks=callbacks, **kwargs)
+        result_dict = {}
+        for key in self.output_keys():
+            result_dict[key] = output_object.get_data(key)
+        return result_dict
+
     def as_langchain_tool(self):
         """Convert to LangChain tool."""
         from langchain.agents.tools import Tool
@@ -239,6 +266,25 @@ class Agent(ComponentBase, ABC):
             description=self.agent_model.info.get("description") + args_description
         )
 
+    async def async_as_langchain_tool(self):
+        """Convert to LangChain tool."""
+        from langchain.agents.tools import Tool
+        format_dict = {}
+        for key in self.input_keys():
+            format_dict.setdefault(key, "input val")
+        format_str = json.dumps(format_dict)
+
+        args_description = f"""
+        to use this tool,your input must be a json string,must contain all keys of {self.input_keys()},
+        and the value of the key must be a json string,the format of the json string is as follows:
+        ```{format_str}```
+        """
+        return Tool(
+            name=self.agent_model.info.get("name"),
+            func=self.async_langchain_run,
+            description=self.agent_model.info.get("description") + args_description
+        )
+
     def process_llm(self, **kwargs) -> LLM:
         llm_name = kwargs.get('llm_name') or self.agent_model.profile.get('llm_model', {}).get('name')
         llm: LLM = LLMManager().get_instance_obj(llm_name)
@@ -338,6 +384,19 @@ class Agent(ComponentBase, ABC):
             tool_results.append(str(tool.run(**tool_input)))
         return "\n\n".join(tool_results)
 
+    async def async_invoke_tools(self, input_object: InputObject, **kwargs) -> str:
+        tool_names = kwargs.get('tool_names') or self.agent_model.action.get('tool', [])
+        if not tool_names:
+            return ''
+        tool_results: list = list()
+        for tool_name in tool_names:
+            tool: Tool = ToolManager().get_instance_obj(tool_name)
+            if tool is None:
+                continue
+            tool_input = {key: input_object.get_data(key) for key in tool.input_keys}
+            tool_results.append(await tool.async_run(**tool_input))
+        return "\n\n".join(tool_results)
+
     def invoke_knowledge(self, query_str: str, input_object: InputObject, **kwargs) -> str:
         knowledge_names = kwargs.get('knowledge_names') or self.agent_model.action.get('knowledge', [])
         if not knowledge_names or not query_str:

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/agentuniverse/agent/template/agent_template.py

@@ -99,6 +99,9 @@ class AgentTemplate(Agent, ABC):
     def invoke_tools(self, input_object: InputObject, **kwargs) -> str:
         return super().invoke_tools(input_object=input_object, tool_names=self.tool_names)
 
+    async def async_invoke_tools(self, input_object: InputObject, **kwargs) -> str:
+        return await super().async_invoke_tools(input_object=input_object, tool_names=self.tool_names)
+
     def invoke_knowledge(self, query_str: str, input_object: InputObject, **kwargs) -> str:
         return super().invoke_knowledge(query_str=query_str, input_object=input_object,
                                         knowledge_names=self.knowledge_names)

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/agentuniverse/agent/template/contextual_iteration_agent_template.py

@@ -66,7 +66,7 @@ class ContextualIterationAgentTemplate(AgentTemplate):
         })
         agent_input['chat_history'] = json.dumps(conversation_history, ensure_ascii=False)
 
-        # build interation chain
+        # build iteration chain
         continue_prompt: Prompt = super(AgentTemplate, self).process_prompt(
             agent_input, prompt_version=self.continue_prompt_version)
         continue_chain = continue_prompt.as_langchain() | llm.as_langchain_runnable(
@@ -138,7 +138,7 @@ class ContextualIterationAgentTemplate(AgentTemplate):
         agent_input['chat_history'] = json.dumps(conversation_history,
                                                  ensure_ascii=False)
 
-        # build interation chain
+        # build iteration chain
         continue_prompt: Prompt = super(AgentTemplate, self).process_prompt(
             agent_input, prompt_version=self.continue_prompt_version)
         continue_chain = continue_prompt.as_langchain() | llm.as_langchain_runnable(

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/agentuniverse/agent/template/rag_agent_template.py

@@ -38,7 +38,7 @@ class RagAgentTemplate(AgentTemplate):
 
     async def customized_async_execute(self, input_object: InputObject, agent_input: dict, memory: Memory, llm: LLM,
                                        prompt: Prompt, **kwargs) -> dict:
-        tool_res: str = self.invoke_tools(input_object)
+        tool_res: str = await self.async_invoke_tools(input_object)
         knowledge_res: str = self.invoke_knowledge(agent_input.get('input'), input_object)
         agent_input['background'] = (agent_input['background']
                                      + f"tool_res: {tool_res} \n\n knowledge_res: {knowledge_res}")

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/agentuniverse/agent/template/react_agent_template.py

@@ -80,7 +80,7 @@ class ReActAgentTemplate(AgentTemplate):
                                        llm: LLM, prompt: Prompt, **kwargs) -> dict:
         self.load_memory(memory, agent_input)
         process_llm_token(llm, prompt.as_langchain(), self.agent_model.profile, agent_input)
-        lc_tools: List[LangchainTool] = self._convert_to_langchain_tool()
+        lc_tools: List[LangchainTool] = await self._async_convert_to_langchain_tool()
         agent = self.create_react_agent(llm.as_langchain(), lc_tools, prompt.as_langchain(),
                                         stop_sequence=self.stop_sequence,
                                         bind_params=self.agent_model.llm_params())
@@ -159,6 +159,22 @@ class ReActAgentTemplate(AgentTemplate):
                 lc_tools.append(agent.as_langchain_tool())
         return lc_tools
 
+    async def _async_convert_to_langchain_tool(self) -> list[LangchainTool]:
+        lc_tools = []
+        if self.tool_names:
+            for tool_name in self.tool_names:
+                tool: Tool = ToolManager().get_instance_obj(tool_name)
+                lc_tools.append(await tool.async_as_langchain())
+        if self.knowledge_names:
+            for knowledge_name in self.knowledge_names:
+                knowledge: Knowledge = KnowledgeManager().get_instance_obj(knowledge_name)
+                lc_tools.append(await knowledge.async_as_langchain_tool())
+        if self.agent_names:
+            for agent_name in self.agent_names:
+                agent: Agent = AgentManager().get_instance_obj(agent_name)
+                lc_tools.append(await agent.async_as_langchain_tool())
+        return lc_tools
+
     def _get_run_config(self, input_object: InputObject) -> RunnableConfig:
         config = RunnableConfig()
         callbacks = []

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/agentuniverse/agent_serve/service_instance.py

@@ -15,7 +15,7 @@ class ServiceInstance(object):
 
     def __init__(self, service_code: str):
         """Initialize a service instance. Raise an ServiceNotFoundError when
-        service code can't be found by servie manager.
+        service code can't be found by service manager.
 
         Args:
             service_code (`str`):

{agentuniverse-0.0.17.dev1 → agentuniverse-0.0.17.dev2}/agentuniverse/agent_serve/web/flask_server.py

@@ -238,9 +238,11 @@ def openai_protocol_chat(model: str, messages: list):
         "messages": messages,
         "stream": stream
     }
+    span = opentracing.tracer.active_span
     if stream:
+        span = opentracing.tracer.active_span
         task = RequestTask(service_run_queue, False, **params)
-        response = Response(timed_generator(task.user_stream_run(), g.start_time), mimetype="text/event-stream")
+        response = Response(timed_generator(task.user_stream_run(), g.start_time,span), mimetype="text/event-stream")
         response.headers['X-Request-ID'] = task.request_id
         return response
     try: