clap-agents 0.3.0__tar.gz → 0.3.1__tar.gz

{clap_agents-0.3.0 → clap_agents-0.3.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: clap-agents
-Version: 0.3.0
+Version: 0.3.1
 Summary: A Python framework for building cognitive agentic patterns including ReAct agents, Multi-Agent Teams, native tool calling, and MCP client integration.
 Project-URL: Homepage, https://github.com/MaitreyaM/CLAP-AGENTS.git
 Project-URL: Repository, https://github.com/MaitreyaM/CLAP-AGENTS.git

{clap_agents-0.3.0 → clap_agents-0.3.1}/examples/google_team.py

@@ -21,6 +21,7 @@ async def run_google_team():
             task_expected_output="A 3-item bullet list.",
             llm_service=google_llm_service,
             model=gemini_model,
+            max_rounds=10
             
         )
 

{clap_agents-0.3.0 → clap_agents-0.3.1}/examples/mcp_team_agent.py

@@ -42,6 +42,7 @@ async def run_agile_team_with_mcp():
             task_description="Analyze the provided context (search results and addition result) and extract the main benefits. Also count the words in the addition result.",
             task_expected_output="A concise bullet-point list summarizing key benefits, and the word count.",
             tools=[count_words],
+            max_rounds=10
         )
         reporter = Agent(
             name="Report_Writer",

{clap_agents-0.3.0 → clap_agents-0.3.1}/examples/weth_test.py

@@ -1,9 +1,7 @@
-# examples/verify_weth_balance.py
 import os
 from dotenv import load_dotenv
 from web3 import Web3
 
-# --- Configuration ---
 load_dotenv()
 WEB3_PROVIDER_URL = os.getenv("WEB3_PROVIDER_URL")
 if not WEB3_PROVIDER_URL:
@@ -14,7 +12,6 @@ ADDRESS_WITH_WETH = "0x785239105435919318a804391305417B62657e05" # An address kn
 AGENT_ADDRESS = Web3().eth.account.from_key(os.getenv("AGENT_PRIVATE_KEY")).address
 
 
-# Standard ERC20 ABI to check balance and decimals
 ERC20_ABI = """
 [
   {"constant":true,"inputs":[{"name":"_owner","type":"address"}],"name":"balanceOf","outputs":[{"name":"balance","type":"uint256"}],"type":"function"},

{clap_agents-0.3.0 → clap_agents-0.3.1}/pyproject.toml

@@ -6,7 +6,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "clap-agents"
-version = "0.3.0"
+version = "0.3.1"
 description = "A Python framework for building cognitive agentic patterns including ReAct agents, Multi-Agent Teams, native tool calling, and MCP client integration."
 readme = "README.md"
 requires-python = ">=3.10"

{clap_agents-0.3.0 → clap_agents-0.3.1}/src/clap/embedding/fastembed_embedding.py

@@ -57,7 +57,7 @@ class FastEmbedEmbeddings(EmbeddingFunctionInterface):
         except Exception as e:
             raise RuntimeError(f"Failed to initialize fastembed model '{self.model_name}': {e}")
 
-    async def __call__(self, input: List[str]) -> List[List[float]]: # Changed to 'input'
+    async def __call__(self, input: List[str]) -> List[List[float]]: 
         if not input: return []
         if not _FASTEMBED_LIB_AVAILABLE: raise RuntimeError("FastEmbed library not available.")
         

{clap_agents-0.3.0 → clap_agents-0.3.1}/src/clap/llm_services/__init__.py

@@ -1,4 +1,3 @@
-# src/clap/llm_services/__init__.py
 from .base import LLMServiceInterface, StandardizedLLMResponse, LLMToolCall
 from .groq_service import GroqService
 from .google_openai_compat_service import GoogleOpenAICompatService

{clap_agents-0.3.0 → clap_agents-0.3.1}/src/clap/llm_services/base.py

@@ -32,7 +32,6 @@ class LLMServiceInterface(abc.ABC):
         messages: List[Dict[str, Any]],
         tools: Optional[List[Dict[str, Any]]] = None,
         tool_choice: str = "auto",
-        # Optional: 
         # temperature: Optional[float] = None,
         # max_tokens: Optional[int] = None,
     ) -> StandardizedLLMResponse:

{clap_agents-0.3.0 → clap_agents-0.3.1}/src/clap/llm_services/ollama_service.py

@@ -64,27 +64,35 @@ class OllamaOpenAICompatService(LLMServiceInterface):
         if not request_model: raise ValueError("Ollama model name not specified.")
         try:
             api_kwargs: Dict[str, Any] = {"messages": messages, "model": request_model}
+
             if tools and tool_choice != "none":
                 api_kwargs["tools"] = tools
                 if isinstance(tool_choice, dict) or tool_choice in ["auto", "required", "none"]: api_kwargs["tool_choice"] = tool_choice
             else: api_kwargs["tools"] = None; api_kwargs["tool_choice"] = None
+
             if temperature is not None: api_kwargs["temperature"] = temperature
             if max_tokens is not None: api_kwargs["max_tokens"] = max_tokens
             api_kwargs = {k: v for k, v in api_kwargs.items() if v is not None}
             # print(f"OllamaService: Sending request to model '{request_model}'")
             response = await self._client.chat.completions.create(**api_kwargs)
+
             message = response.choices[0].message
+
             text_content = message.content
             tool_calls_std: List[LLMToolCall] = []
+
             if message.tool_calls:
                 for tc in message.tool_calls:
                     if tc.id and tc.function and tc.function.name and tc.function.arguments is not None:
                         tool_calls_std.append(LLMToolCall(id=tc.id, function_name=tc.function.name, function_arguments_json_str=tc.function.arguments))
                     else: print(f"{Fore.YELLOW}Warning: Incomplete tool_call from Ollama: {tc}{Fore.RESET}")
+
             return StandardizedLLMResponse(text_content=text_content, tool_calls=tool_calls_std)
+        
         except _OpenAIError_Placeholder_Type as e: # Use placeholder
             err_msg = f"Ollama (OpenAI Compat) API Error: {e}"
             if hasattr(e, 'response') and e.response and hasattr(e.response, 'text'): err_msg += f" - Details: {e.response.text}"
+            
             print(f"{Fore.RED}{err_msg}{Fore.RESET}")
             return StandardizedLLMResponse(text_content=err_msg)
         except Exception as e:

{clap_agents-0.3.0 → clap_agents-0.3.1}/src/clap/mcp_client/client.py

@@ -39,7 +39,7 @@ class MCPClientManager:
         self._connect_locks: Dict[str, asyncio.Lock] = {
              name: asyncio.Lock() for name in server_configs
         }
-        self._manager_lock = asyncio.Lock() # General lock for manager state
+        self._manager_lock = asyncio.Lock() 
 
     async def _ensure_connected(self, server_name: str):
         """

{clap_agents-0.3.0 → clap_agents-0.3.1}/src/clap/multiagent_pattern/agent.py

@@ -74,6 +74,7 @@ class Agent:
         mcp_server_names: Optional[List[str]] = None,
         vector_store: Optional[VectorStoreInterface] = None,
         parallel_tool_calls: bool = True , 
+        **kwargs
         # embedding_function: Optional[EmbeddingFunction] = None,
 
     ):
@@ -86,6 +87,7 @@ class Agent:
         self.local_tools = tools or [] 
 
         self.vector_store = vector_store
+        self.react_agent_kwargs = kwargs
         # self.embedding_function = embedding_function 
 
         llm_service_instance = llm_service or GroqService()
@@ -193,7 +195,7 @@ class Agent:
         self.task_description = original_task_description 
 
         print(f"Agent {self.name}: Running ReactAgent...")
-        raw_output = await self.react_agent.run(user_msg=msg) 
+        raw_output = await self.react_agent.run(user_msg=msg,**self.react_agent_kwargs) 
         output_data = {"output": raw_output}
 
         print(f"Agent {self.name}: Passing context to {len(self.dependents)} dependents...")

{clap_agents-0.3.0 → clap_agents-0.3.1}/src/clap/tool_pattern/tool.py

@@ -16,9 +16,11 @@ def get_fn_signature(fn: Callable) -> dict:
     sig = inspect.signature(fn)
     for name, type_hint in fn.__annotations__.items():
         if name == "return": continue
+        
         param_type_name = getattr(type_hint, "__name__", str(type_hint))
         schema_type = type_mapping.get(param_type_name.lower(), "string") 
         parameters["properties"][name] = {"type": schema_type}
+
         if sig.parameters[name].default is inspect.Parameter.empty:
             parameters["required"].append(name)
     if not parameters.get("required"):