Update agent.py

agent.py

@@ -21,6 +21,14 @@ from langchain.embeddings.base import Embeddings
 from typing import List
 import numpy as np
 
+import json
+import uuid
+from langchain.embeddings import HuggingFaceEmbeddings
+from langchain.vectorstores import FAISS
+from langchain.schema import Document
+from langchain.tools.retriever import create_retriever_tool
+from sentence_transformers import SentenceTransformer
+
 load_dotenv()
 
 @tool
@@ -125,52 +133,47 @@ with open("system_prompt.txt", "r", encoding="utf-8") as f:
 # System message
 sys_msg = SystemMessage(content=system_prompt)
 
-# Custom embedding class
-
-#from langchain_huggingface import HuggingFaceEmbeddings
-
-
-
-# Initialize SentenceTransformer and set max_seq_length
-sentence_transformer = SentenceTransformer("sentence-transformers/all-mpnet-base-v2")
-sentence_transformer.max_seq_length = 512  # Set max sequence length
-
-# Initialize embeddings with the model name (dim=768)
-embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-mpnet-base-v2")
-
-#from sentence_transformers import SentenceTransformer
 
-model = SentenceTransformer("sentence-transformers/all-mpnet-base-v2")
-query = "What is vector search?"
-query_embedding = model.encode(query)
 
-print("Embedding Length:", len(query_embedding))  # Ensure it's 768
+# -------------------------------
+# Step 1: Load metadata.jsonl (max 165 docs)
+# -------------------------------
+docs = []
+with open("metadata.jsonl", "r", encoding="utf-8") as f:
+    for i, line in enumerate(f):
+        if i >= 165:
+            break
+        data = json.loads(line)
+        content = data.pop("content", "").strip()
+        if not content:
+            continue  # skip empty
+        data["id"] = str(uuid.uuid4())  # ensure each doc has unique ID
+        docs.append(Document(page_content=content, metadata=data))
+
+# -------------------------------
+# Step 2: Set up Embeddings + FAISS
+# -------------------------------
+embedding_model = HuggingFaceEmbeddings(model_name="sentence-transformers/all-mpnet-base-v2")
+vector_store = FAISS.from_documents(docs, embedding_model)
+
+# Save FAISS index locally (optional)
+vector_store.save_local("faiss_index")
+
+# -------------------------------
+# Step 3: Create Retriever Tool
+# -------------------------------
+retriever = vector_store.as_retriever()
 
-
-
-
-# Initialize Supabase client
-supabase: Client = create_client(
-    os.environ.get("SUPABASE_URL"),
-    os.environ.get("SUPABASE_SERVICE_KEY")
-)
-
-# Initialize Supabase vector store
-vector_store = SupabaseVectorStore(
-    client=supabase,
-    embedding=embeddings,
-    table_name="documents",
-    query_name="match_documents_langchain"
-    )
-    
 # Create retriever tool
-retriever_tool = create_retriever_tool(
-    retriever=vector_store.as_retriever(),
-    name="Question Search",
-    description="A tool to retrieve similar questions from a vector store."
+question_retriever_tool = create_retriever_tool(
+    retriever=retriever,
+    name="Question_Search",
+    description="Use this tool to retrieve documents related to a user's question."
 )
 
 
+
+
 tools = [
     multiply,
     add,