RAG CLI Tool

A command-line interface for Retrieval-Augmented Generation (RAG) that enables querying documents using AI-powered search and answer generation.

Overview

The RAG CLI Tool provides a robust solution for document-based question answering by combining the power of large language models with semantic search capabilities. It processes web documents, creates vector embeddings, and generates contextual answers based on retrieved information.

Features

• Document Retrieval: Load and process documents from web URLs with configurable parsing
• AI-Powered Answers: Generate contextual answers using OpenAI's GPT models
• Interactive Mode: Chat-like interface for multiple queries with command history
• Customizable Parameters: Adjust chunk sizes, URLs, and other processing parameters
• Verbose Output: Detailed information about the retrieval and processing pipeline
• Environment Management: Simple configuration through environment variables
• Error Handling: Comprehensive error checking and user-friendly error messages

Installation

Prerequisites

• Python 3.8 or higher
• OpenAI API key
• Internet connection for document loading

Setup Instructions

1. Clone or download the project files

2. Install dependencies:

   pip install -r requirements.txt

3. Configure environment variables: Create a .env file in the project directory with:

   OPENAI_API_KEY=your_openai_api_key_here
   USER_AGENT=your_user_agent_string_here

Usage

Basic Commands

Environment verification:

python rag_cli.py --check-env

Single question query:

python rag_cli.py --question "What is Task Decomposition?"

Interactive mode (default behavior):

python rag_cli.py --interactive
# or simply:
python rag_cli.py

Advanced Configuration

Custom document sources:

python rag_cli.py --urls "https://example.com/article1" "https://example.com/article2" --interactive

Optimized chunk settings:

python rag_cli.py --chunk-size 500 --chunk-overlap 100 --question "What is AI?"

Detailed processing information:

python rag_cli.py --verbose --question "What is machine learning?"

Interactive Mode Commands

When running in interactive mode, the following commands are available:

• Enter any question to receive an AI-generated answer
• help - Display available commands and usage information
• quit, exit, or q - Terminate the application

Examples

Example 1: Basic Query Processing

$ python rag_cli.py --question "What is Task Decomposition?"
Setting up RAG system...
   URLs: ['https://lilianweng.github.io/posts/2023-06-23-agent/']
   Chunk size: 1000
   Chunk overlap: 200
Loading documents from URLs...
   Loaded 1 documents
Splitting documents into chunks...
   Split into 15 chunks
Creating vector store...
RAG system ready!

Question: What is Task Decomposition?
Searching for relevant context...

Answer:
Task decomposition is a technique used in AI and machine learning where complex tasks are broken down into smaller, more manageable subtasks...

Example 2: Interactive Session

$ python rag_cli.py --interactive
Setting up RAG system...
   URLs: ['https://lilianweng.github.io/posts/2023-06-23-agent/']
   Chunk size: 1000
   Chunk overlap: 200
Loading documents from URLs...
   Loaded 1 documents
Splitting documents into chunks...
   Split into 15 chunks
Creating vector store...
RAG system ready!

Interactive RAG Query Mode
   Type 'quit' or 'exit' to stop
   Type 'help' for available commands

Enter your question: What is an AI agent?
Searching for relevant context...

Answer:
An AI agent is an autonomous system that can perceive its environment, make decisions, and take actions to achieve specific goals...

Enter your question: quit
Goodbye!

Configuration

Environment Variables

Variable	Description	Required
OPENAI_API_KEY	Your OpenAI API key for model access	Yes
USER_AGENT	User agent string for web scraping	Yes

Command Line Options

Option	Short	Description	Default
--question	-q	Single question to process	None
--interactive	-i	Enable interactive mode	False
--urls	-u	Document URLs to process	Lilian Weng's agent post
--chunk-size		Text chunk size in characters	1000
--chunk-overlap		Overlap between chunks	200
--verbose	-v	Enable detailed output	False
--check-env		Verify environment configuration	False

Technical Architecture

Processing Pipeline

1. Document Loading: Web scraping with BeautifulSoup for content extraction
2. Text Processing: Recursive character-based text splitting with configurable parameters
3. Vector Embedding: OpenAI's text-embedding-3-large model for semantic representation
4. Similarity Search: In-memory vector store for efficient document retrieval
5. Answer Generation: GPT-5-mini model with retrieved context for response generation

Technology Stack

• Language Model: GPT-5-mini (OpenAI)
• Embedding Model: text-embedding-3-large (OpenAI)
• Vector Store: InMemoryVectorStore (LangChain)
• Orchestration: LangGraph for workflow management
• Text Processing: RecursiveCharacterTextSplitter
• Web Scraping: BeautifulSoup with custom selectors

Performance Considerations

• Memory Usage: In-memory vector store requires sufficient RAM for large document collections
• API Costs: OpenAI API usage incurs costs based on token consumption
• Processing Time: Initial setup includes document loading and embedding generation
• Network Dependency: Requires internet access for document retrieval and API calls

Troubleshooting

Common Issues

1. Missing API Key: Ensure OPENAI_API_KEY is properly set in your environment
2. Missing User Agent: Configure USER_AGENT for web scraping compliance
3. Import Errors: Install all dependencies using the provided requirements.txt
4. Network Connectivity: Verify internet access for document loading and API calls
5. Memory Limitations: Consider reducing chunk size for large document collections

Error Resolution

Environment verification:

python rag_cli.py --check-env

Dependency installation:

pip install -r requirements.txt

Help documentation:

python rag_cli.py --help

Development

Project Structure

RAG-Test/
├── rag_cli.py          # Main CLI application
├── requirements.txt    # Python dependencies
├── setup.py           # Package configuration
├── README.md          # Documentation
├── rag-cli.bat        # Windows execution script
├── RAG-Test.py        # Original implementation
└── RAG-Test.ipynb     # Jupyter notebook version

Contributing

1. Fork the repository
2. Create a feature branch
3. Implement your changes
4. Add appropriate tests
5. Submit a pull request

License

This project is licensed under the MIT License. See the LICENSE file for details.

Support

For issues, questions, or contributions, please refer to the project documentation or create an issue in the repository.