refactor: Update local RAG agent to use AgentOS and enhance knowledge management

- Replaced PDFUrlKnowledgeBase with Knowledge class for improved document management.
- Updated to use AgentOS for the application interface instead of Playground.
- Adjusted README to reflect changes in the interface and knowledge base loading process.
- Specified minimum version for the 'agno' package in requirements.txt.

rag_tutorials/local_rag_agent/README.md

@@ -1,13 +1,14 @@
 ## 🦙 Local RAG Agent with Llama 3.2
-This application implements a Retrieval-Augmented Generation (RAG) system using Llama 3.2 via Ollama, with Qdrant as the vector database.
+This application implements a Retrieval-Augmented Generation (RAG) system using Llama 3.2 via Ollama, with Qdrant as the vector database. Built with Agno v2.0.
 
 
 ### Features
 - Fully local RAG implementation
 - Powered by Llama 3.2 through Ollama
 - Vector search using Qdrant
-- Interactive playground interface
+- Interactive AgentOS interface
 - No external API dependencies
+- Uses Agno v2.0 Knowledge class for document management
 
 ### How to get Started?
 
@@ -36,11 +37,15 @@ ollama pull llama3.2
 ollama pull openhermes
 ```
 
-4. Run the AI RAG Agent 
+5. Run the AI RAG Agent 
 ```bash
 python local_rag_agent.py
 ```
 
-5. Open your web browser and navigate to the URL provided in the console output to interact with the RAG agent through the playground interface.
+6. Open your web browser and navigate to the URL provided in the console output (typically `http://localhost:7777`) to interact with the RAG agent through the AgentOS interface.
+
+### Note
+- The knowledge base loads a Thai Recipes PDF on the first run. You can comment out the `knowledge_base.add_content()` line after the first run to avoid reloading.
+- The AgentOS interface provides a web-based UI for interacting with your agent.
 
 

rag_tutorials/local_rag_agent/local_rag_agent.py

@@ -1,10 +1,10 @@
 # Import necessary libraries
 from agno.agent import Agent
 from agno.models.ollama import Ollama
-from agno.knowledge.pdf_url import PDFUrlKnowledgeBase
+from agno.knowledge.knowledge import Knowledge
 from agno.vectordb.qdrant import Qdrant
-from agno.embedder.ollama import OllamaEmbedder
+from agno.knowledge.embedder.ollama import OllamaEmbedder
-from agno.playground import Playground, serve_playground_app
+from agno.os import AgentOS
 
 # Define the collection name for the vector database
 collection_name = "thai-recipe-index"
@@ -16,14 +16,15 @@ vector_db = Qdrant(
     embedder=OllamaEmbedder()
 )
 
-# Define the knowledge base with the specified PDF URL
+# Define the knowledge base
-knowledge_base = PDFUrlKnowledgeBase(
+knowledge_base = Knowledge(
-    urls=["https://phi-public.s3.amazonaws.com/recipes/ThaiRecipes.pdf"],
     vector_db=vector_db,
 )
 
-# Load the knowledge base, comment out after the first run to avoid reloading
+# Add content to the knowledge base, comment out after the first run to avoid reloading
-knowledge_base.load(recreate=True, upsert=True)
+knowledge_base.add_content(
+    url="https://phi-public.s3.amazonaws.com/recipes/ThaiRecipes.pdf"
+)
 
 # Create the Agent using Ollama's llama3.2 model and the knowledge base
 agent = Agent(
@@ -33,8 +34,9 @@ agent = Agent(
 )
 
 # UI for RAG agent
-app = Playground(agents=[agent]).get_app()
+agent_os = AgentOS(agents=[agent])
+app = agent_os.get_app()
 
-# Run the Playground app
+# Run the AgentOS app
 if __name__ == "__main__":
-    serve_playground_app("local_rag_agent:app", reload=True)
+    agent_os.serve(app="local_rag_agent:app", reload=True)

rag_tutorials/local_rag_agent/requirements.txt

@@ -1,7 +1,4 @@
-agno
+agno>=2.2.10
 qdrant-client
 ollama
 pypdf 
-openai
-fastapi
-uvicorn