fix: langchain buffermemory try

langchain_mcp_client.py

@@ -18,9 +18,10 @@ from langchain_google_genai import ChatGoogleGenerativeAI
 from langchain_core.prompts import PromptTemplate
 
 
-
 logger = logging.getLogger(__name__)
 load_dotenv()
+
+
 async def lc_mcp_exec(request: str, history) -> str:
     """
     Execute the full PostgreSQL MCP pipeline: load summary, connect session,
@@ -45,7 +46,9 @@ async def lc_mcp_exec(request: str, history) -> str:
                 api_key=api_key,
                 temperature=0.5,
             )
-    print(type(llm))
+
+
+
     async with stdio_client(server_params) as (read, write):
         async with ClientSession(read, write) as session:
             await session.initialize()
@@ -55,15 +58,15 @@ async def lc_mcp_exec(request: str, history) -> str:
             intent = classify_intent(request)
 
             memory = ConversationBufferMemory(return_messages=True)
-            agent = create_react_agent(llm, tools)
-
             prompt = await build_prompt(session, intent, request, tools, table_summary)
+
+            agent = create_react_agent(llm, tools)
             agent_executor = AgentExecutor(agent=agent, tools=tools, memory=memory, verbose=True)
-            config = {"configurable": {"thread_id": "conversation_123"}}
+
             result = await agent_executor.ainvoke(
-                {"messages": prompt},
-                config
+                {"messages": prompt}
             )
+
             response = result.get("output", "No response generated")
 
             return response
@@ -84,6 +87,8 @@ def get_server_params() -> StdioServerParameters:
 
 async def load_and_enrich_tools(session: ClientSession):
     tools = await load_mcp_tools(session)
+    print("Loaded tools:", tools)
+    print("Tool types:", [type(t) for t in tools])
     return tools
 
 async def build_prompt(session, intent, request, tools, summary):
@@ -97,8 +102,10 @@ async def build_prompt(session, intent, request, tools, summary):
         )
     else:
         template = conversation_prompt.contents[0].text
-        #tools_str = "\n".join([f"- {tool.name}: {tool.description}" for tool in tools])
-        #return template.format(  new_request=request,  tools=tools_str, descriptions=summary  )
-        prompt = PromptTemplate.from_template(template)
-        return prompt
+        tools_str = "\n".join([f"- {tool.name}: {tool.description}" for tool in tools])
+        return template.format(  new_request=request,  tool_names=tools_str, descriptions=summary  )
+        #prompt = PromptTemplate.from_template(template)
+        #return prompt
+
+
 

postgre_mcp_server.py

@@ -133,7 +133,6 @@ async def base_prompt_query() -> str:
 
                     You can use the following FastMCP tools. These allow you to create **read-only** queries, such as `SELECT`, `COUNT`, or queries with `GROUP BY`, `ORDER BY`, and similar clauses. You may chain tools together to gather the necessary information before generating your SQL query.
 
-                    {tools}
                     {tool_names}
 
                     ---
@@ -179,9 +178,7 @@ async def base_prompt_query() -> str:
 
                     Please fulfill the following request based on the above context:
 
-                    {input}
-                    
-                    Thought:{agent_scratchpad}
+                    {new_request}
                     """
 
     return base_prompt