Discord Bot with LLM Integration

This project is a Discord bot that uses Retrieval-Augmented Generation (RAG) from EdenAI as a chatbot to respond to users' questions. Built with Python, it leverages the discord.py library for Discord interactions and integrates with a Large Language Model (LLM) via EdenAI's API to deliver context-aware, intelligent responses.

Features

• Ping Command: Check the bot's latency with the !ping command.
• Ask Command: Query the LLM with the !ask <query> command to get intelligent responses.
• Phoenix Integration: Traces and monitors LLM interactions using the Phoenix library.

Project Structure

.env.example
main.py
phoenix_setup.py
utils.py

• main.py: The main entry point for the Discord bot.
• phoenix_setup.py: Configures Phoenix for tracing and monitoring.
• utils.py: Contains utility functions, including the ask_llm function for interacting with the LLM API.
• .env.example: Example environment variables file.

Prerequisites

• A Discord bot token
• EDENAI API keys for the LLM
• API TOKEN FOR PHEONIX

Setup

1. Clone the repository:

2. Install dependencies:

3. Create a .env file by copying .env.example:

   cp .env.example .env

4. Fill in the .env file with your credentials:

   DISCORD_BOT_TOKEN=<your-discord-bot-token>
   RAG_PROJECT_ID=<your-rag-project-id>
   API_KEY=<your-api-key>
   PHOENIX_API_TOKEN=<your-phoenix-api-token>
   PHOENIX_COLLECTOR_ENDPOINT=<your-phoenix-collector-endpoint>
   PHOENIX_CLIENT_HEADERS=api_key=<your-phoenix-client-api-key>

Running the Bot

Start the bot by running the following command:

python main.py

Usage

• Ping Command: Type !ping in a Discord channel to check the bot's latency.
• Ask Command: Type !ask <your-query> in a Discord channel to get a response from the LLM.

Environment Variables

The bot uses the following environment variables:

• DISCORD_BOT_TOKEN: Your Discord bot token.
• RAG_PROJECT_ID: The project ID for the LLM API.
• API_KEY: The API key for the LLM service.
• PHOENIX_API_TOKEN: The API token for Phoenix.
• PHOENIX_COLLECTOR_ENDPOINT: The endpoint for Phoenix traces.
• PHOENIX_CLIENT_HEADERS: Additional headers for Phoenix.