Create agent.py

agent.py

@@ -0,0 +1,67 @@
+from langchain import hub
+from langchain.agents import AgentExecutor, create_react_agent
+from langchain_openai import OpenAI
+from langchain_community.tools import DuckDuckGoSearchResults
+
+class ReActAgent:
+  """
+  A LangChain agent class with conversation history for contextual processing.
+  """
+
+  def __init__(self):
+    """
+    Initializes the agent with default tools, OpenAI LLM, and an empty history.
+    """
+    self.tools = [DuckDuckGoSearchResults()]
+    self.prompt_template = '''
+    Answer the following questions as best you can, considering the conversation history:
+
+    {history}
+
+    You have access to the following tools:
+
+    {tools}
+
+    Use the following format:
+
+    Question: the input question you must answer
+    Thought: you should always think about what to do
+    Action: the action to take (use history first, then tools if needed)
+    Action Input: the input to the action
+    Observation: the result of the action
+    ... (this Thought/Action/Action Input/Observation can repeat N times)
+    Thought: I now know the final answer
+    Final Answer: the final answer to the original input question
+
+    Begin!
+
+    Question: {input}
+    Thought:{agent_scratchpad}
+
+    '''
+    self.llm = OpenAI()
+    
+  def update_history(self, question, answer):
+    """
+    Updates the conversation history with the latest question and answer.
+    """
+    self.history.append({"question": question, "answer": answer})
+
+  def create_agent(self, history):
+    """
+    Creates a ReAct agent based on the defined prompt, LLM, and history.
+    """
+    history_str = "\n".join([f"{i['question']}: {i['answer']}" for i in history])
+    prompt = self.prompt_template.format(history=history_str, tools=",".join(t.__name__ for t in self.tools))
+    agent = create_react_agent(self.llm, self.tools, prompt)
+    return agent
+
+  def run(self, question,history=[], verbose=True):
+    """
+    Executes the agent with the provided question, verbosity option, and updates history.
+    """
+    agent = self.create_agent(history)
+    agent_executor = AgentExecutor(agent=agent, tools=self.tools, verbose=verbose)
+    answer = agent_executor.invoke({"input": question, "history": history})
+    return answer
+