shivonai-mcp

# tested on openai models and bedrock claude models # work with both types of models models. # #openai use # from langchain_openai import ChatOpenAI # from langchain.agents import initialize_agent, AgentType # from shivonai.lyra import langchain_toolkit # # Replace with your actual MCP server details # auth_token = "shivonai_auth_token" # # Get LangChain tools # tools = langchain_toolkit(auth_token) # # Print available tools # print(f"Available tools: {[tool.name for tool in tools]}") # # Initialize LangChain agent with tools # llm = ChatOpenAI( # temperature=0, # model_name="gpt-4-turbo", # openai_api_key="openai-api-key" # ) # agent = initialize_agent( # tools=tools, # llm=llm, # agent=AgentType.ZERO_SHOT_REACT_DESCRIPTION, # verbose=True # ) # # Try running the agent with a simple task # try: # result = agent.run("what listing I have?") # print(f"Result: {result}") # except Exception as e: # print(f"Error: {e}") # bedrock use from langchain_aws import ChatBedrock from langchain.agents import initialize_agent, AgentType from shivonai.lyra import langchain_toolkit import os os.environ["AWS_ACCESS_KEY_ID"] = "bedrock_access_key" os.environ["AWS_SECRET_ACCESS_KEY"] = "bedrock_secrate_access_key" # Replace with your actual MCP server details auth_token = "Shivonai_auth_token" # Get LangChain tools tools = langchain_toolkit(auth_token) # Print available tools print(f"Available tools: {[tool.name for tool in tools]}") # Initialize LangChain agent with Claude Sonnet via AWS Bedrock llm = ChatBedrock( model_id="anthropic.claude-3-sonnet-20240229-v1:0", # Claude Sonnet model ID region_name="bedrock_region", # Replace with your AWS region temperature=0 ) agent = initialize_agent( tools=tools, llm=llm, agent=AgentType.ZERO_SHOT_REACT_DESCRIPTION, verbose=True ) # Try running the agent with a simple task try: result = agent.run("what listing I have?") print(f"Result: {result}") except Exception as e: print(f"Error: {e}")