feat: a functional RAG agent that retrive data from inmemory KB acording to the prompt context

agents/BookRetriverAgent.py

@@ -4,21 +4,22 @@ from langchain_huggingface import HuggingFaceEndpoint, ChatHuggingFace
 import os
 
 
-hf_model = HuggingFaceEndpoint(
-    repo_id="Qwen/Qwen2.5-Coder-7B-Instruct",
-    task="text-generation",
-    provider="auto",
-    huggingfacehub_api_token=os.getenv("HF_TOKEN")
-)
-llm = ChatHuggingFace(llm=hf_model)
+def get_book_retriver_agent():
+    hf_model = HuggingFaceEndpoint(
+        repo_id="Qwen/Qwen3-30B-A3B-Instruct-2507",
+        task="text-generation",
+        provider="auto",
+        huggingfacehub_api_token=os.getenv("HF_TOKEN")
+    )
+    llm = ChatHuggingFace(llm=hf_model)
 
-tools = [retrieve_book_context]
-prompt = (
-    "You have access to a tool that retrieves context from books."
-    "Use the tool to help answer user queries."
-)
-agent = create_agent(
-    llm,
-    tools,
-    system_prompt=prompt
-)
+    tools = [retrieve_book_context]
+    prompt = (
+        "You have access to a tool that retrieves context from books."
+        "Use the tool to help answer user queries."
+    )
+    return create_agent(
+        llm,
+        tools,
+        system_prompt=prompt
+    )

agents/__init__.py

@@ -1 +1 @@
-from .BookRetriverAgent import *
+from .BookRetriverAgent import get_book_retriver_agent

app.py

@@ -4,6 +4,7 @@ from langchain.agents import create_agent
 from langchain_huggingface import HuggingFaceEndpoint, ChatHuggingFace
 from langgraph.checkpoint.memory import InMemorySaver
 from dotenv import load_dotenv
+from tools import retrieve_book_context_rag_agent
 
 
 class GradioAgent:
@@ -27,12 +28,13 @@ class GradioAgent:
             repo_id="Qwen/Qwen3-30B-A3B-Instruct-2507",
             task="text-generation",
             provider="auto",
-            huggingfacehub_api_token=os.getenv("HF_TOKEN")
+            huggingfacehub_api_token=os.getenv("HF_TOKEN"),
+            do_sample=False
         )
         llm = ChatHuggingFace(llm=hf_model)
 
         return create_agent(
-            tools=[],
+            tools=[retrieve_book_context_rag_agent],
             model=llm,
             checkpointer=InMemorySaver(),
             system_prompt="You are a helpful and usefull assistant."

rag/BooksRag.py

@@ -21,7 +21,12 @@ def get_books_vector_store():
         os.getenv("AWS_SECRET_ACCESS_KEY"),
         "us-east-1"
     )
-    all_splits = text_splitter.split_documents(docs)
-    vector_store.add_documents(documents=all_splits)
+    print("retrieved docs")
+    for doc in docs:
+        all_splits = text_splitter.split_documents(doc)
+        print("show")
+        vector_store.add_documents(documents=all_splits)
+        print("pri")
+
 
     return vector_store

requirements.txt

@@ -7,6 +7,7 @@ annotated-types==0.7.0
 antlr4-python3-runtime==4.9.3
 anyio==4.11.0
 attrs==25.4.0
+audioop-lts==0.2.2
 Authlib==1.6.5
 backoff==2.2.1
 bcrypt==5.0.0

tools/BookRetriverRagAgent.py

@@ -0,0 +1,12 @@
+from langchain.tools import tool
+from agents import get_book_retriver_agent
+
+
+@tool
+def retrieve_book_context_rag_agent(request: str):
+    """Used to retrieve computer science related books context and answer the request."""
+    agent = get_book_retriver_agent()
+    result = agent.invoke({
+        "messages": [{"role": "user", "content": request}]
+    })
+    return result["messages"][-1].text

tools/__init__.py

@@ -1 +1,2 @@
 from .RetriveBooksDataTool import retrieve_book_context
+from .BookRetriverRagAgent import retrieve_book_context_rag_agent

util/aws/S3Access.py

@@ -5,13 +5,13 @@ from langchain_community.document_loaders import S3FileLoader
 def retrieve_s3_data_as_documents(
         bucket_name: str,
         prefix: str,
-        access_key: str,
+        access_id: str,
         secret_key: str,
         region: str
 ):
     s3 = boto3.client(
         's3',
-        aws_access_key_id=access_key,
+        aws_access_key_id=access_id,
         aws_secret_access_key=secret_key,
         region_name=region
     )
@@ -26,8 +26,9 @@ def retrieve_s3_data_as_documents(
         loader = S3FileLoader(
             bucket=bucket_name,
             key=key,
-            aws_access_key_id=access_key,
-            aws_secret_access_key=secret_key
+            aws_access_key_id=access_id,
+            aws_secret_access_key=secret_key,
+            strategy="fast"
         )
         doc = loader.load()
         data.append(doc)