ChanakyaX

ChanakyaX is a Discord bot that allows users to interact with a local Large Language Model (LLM) backend (default: deepseek-r1:7b) via slash commands. ChanakyaX manages a queue for prompt processing and creates threads for each user prompt, streaming responses in real time.

---

🚀 Features

• Slash command /prompt to send prompts to the LLM
• Queue system to manage multiple user requests
• Automatic thread creation for each prompt
• Real-time streaming of LLM responses
• Built with discord.js and TypeScript

---

🖥️ 1. Install and Run Ollama (LLM Backend)

ChanakyaX requires a local LLM backend. We recommend Ollama for Windows.

Step 1: Download and Install Ollama

• Go to the Ollama Windows Download Page
• Download the Windows installer (.exe file)
• Run the installer and follow the prompts

Step 2: Run Ollama and Download a Model

Open Command Prompt or PowerShell and run:

ollama run llama2

• This will download and start the Llama 2 model. You can also use other models (see Ollama Models).
• Keep this terminal open and running while you use ChanakyaX.

Note: ChanakyaX expects Ollama's API to be available at http://localhost:11434/api/generate.

---

🤖 2. Create a Discord Bot (Discord Developer Portal)

1. Go to the Discord Developer Portal
2. Click New Application and give it a name (e.g., ChanakyaX)
3. Go to Bot in the sidebar, then click Add Bot
4. Under Token, click Reset Token and copy your Bot Token
5. Go to OAuth2 > General and copy your Client ID
6. Under OAuth2 > URL Generator, select bot and applications.commands scopes, and set permissions (e.g., Send Messages, Create Public Threads)
7. Copy the generated URL, open it in your browser, and invite ChanakyaX to your server

---

⚙️ 3. Configure Environment Variables

Create a .env file in the root directory of this project:

DISCORD_LLM_BOT_TOKEN=your-discord-bot-token-here
DISCORD_LLM_BOT_CLIENT_ID=your-discord-client-id-here

Replace the values with your actual Bot Token and Client ID from the Discord Developer Portal.

---

🛠️ 4. Install and Build ChanakyaX

1. Clone the repository:

   git clone <your-repo-url>
   cd discord-bot-llm

2. Install dependencies:

   npm install

3. Build the project:

   npx tsc

---

▶️ 5. Run ChanakyaX

1. Start your LLM backend (Ollama) as described above
2. Start ChanakyaX:

   node dist/bot.js

   Or, for development with TypeScript:

   npx ts-node src/bot.ts

3. ChanakyaX will deploy slash commands automatically on startup
4. Use the /prompt command in your Discord server:
   • /prompt input: <your message>
   • ChanakyaX will create a thread and stream the LLM's response there

---

🧩 How It Works

• ChanakyaX listens for the /prompt command
• User prompts are queued and processed in order (up to 3 concurrently)
• For each prompt, a new thread is created in the channel
• The bot sends the prompt to the LLM backend and streams the response in real time, updating the thread message(s)
• The queue system ensures fair processing and rate limit compliance

---

🛠️ Customization

• LLM Model: The default model is set to deepseek-r1:7b in src/queue/queue.ts. You can change this as needed
• LLM Backend URL: The backend URL is hardcoded as http://localhost:11434/api/generate in src/queue/queue.ts. Update this if your backend runs elsewhere

---

👩‍💻 Development

• TypeScript source is in src/, compiled output in dist/
• Add new commands in src/commands/ following the structure of the prompt command
• Type definitions for custom Discord.js extensions are in src/types/discord.d.ts

---

📄 License

ISC

---

This project is not affiliated with Discord or any LLM provider. Use responsibly and ensure compliance with Discord's terms of service.