MistralBot

A sophisticated IRC bot powered by Mistral AI that provides intelligent chat assistance and code help. MistralBot brings the power of large language models to IRC channels through a simple interface.

IRC Server Details

Join us on MansionNET IRC to chat with us, test the bot, and play some trivia!

🌐 Server: irc.inthemansion.com
🔒 Port: 6697 (SSL)
📝 Channel: #opers, #general, #welcome, #devs (and many others)

Features

• AI-powered chat assistance using Mistral AI
• Code explanation and generation
• Built-in rate limiting and user quotas
• SSL/TLS support for secure IRC connections
• Customizable prompt templates
• Smart message handling for IRC message length limits
• Debug logging

Requirements

• Python 3.8 or higher
• Mistral AI API key (Get one here)
• requests and python-dotenv libraries
• SSL-enabled IRC server

Installation

1. Clone the repository:

git clone https://github.com/MansionNET/mistralbot.git
cd mistralbot

2. Create and activate a virtual environment (recommended):

python -m venv venv
source venv/bin/activate  # On Windows use: venv\Scripts\activate

3. Install dependencies:

pip install -r requirements.txt

4. Set up your environment variables:

cp .env.example .env

5. Edit .env and add your Mistral AI API key:

MISTRAL_API_KEY=your_api_key_here

Usage

Start the bot:

python mistralbot.py

Available Commands

In any channel where the bot is present:

• !ask <question> - Ask a general question to the AI

  • Example: !ask What is the capital of France?
  • Example: !ask Can you explain quantum computing?

• !code <question> - Get programming help with code examples

  • Example: !code How do I read a file in Python?
  • Example: !code Write a simple web server

• !help - Display available commands and usage information

Configuration

The bot can be configured through environment variables or by editing the code:

Environment Variables

• MISTRAL_API_KEY (required): Your Mistral AI API key

Code Configuration

• IRC Server settings

  • Server address
  • Port
  • Nickname
  • Channels to join

• Rate Limiting

  • Requests per minute (default: 10)
  • Requests per day (default: 1000)
  • Per-user daily limit (default: 50)

• Prompt Templates

  • Default conversation template
  • Code explanation template
  • Custom templates can be added

Rate Limits

To prevent abuse and ensure fair usage, MistralBot implements several rate limits:

• Global limits:

  • 10 requests per minute
  • 1000 requests per day

• Per-user limits:

  • 50 requests per day
  • 3-second cooldown between commands

Prompt Templates

MistralBot uses customizable prompt templates for different types of queries:

prompt_templates = {
    "default": "You are a helpful assistant. Provide clear, direct responses without unnecessary detail. Question: {query}",
    "code": "You are a programming teacher. First give a ONE-SENTENCE explanation. Then after 'CODE:', show simple, practical code examples with minimal comments. Keep both explanation and code concise. Question: {query}",
    "explain": "Explain this concept in 2-3 short, clear sentences: {query}"
}

You can modify these templates or add new ones to customize the bot's responses.

API Usage

MistralBot uses the Mistral AI API with the following configuration:

• Model: mistral-tiny
• Max tokens: 300
• Temperature: 0.7

Please review Mistral AI's pricing and terms of service before deployment.

Contributing

We welcome contributions! Please read our Contributing Guide for details on:

• Code style
• Pull request process
• Development setup
• Testing requirements

License

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

Security

Never share your Mistral AI API key or commit it to version control. Always use environment variables or secure configuration management.

Project Status

This project is actively maintained. For support:

• Open an issue for bugs or feature requests
• Check existing issues before creating new ones
• Follow the contribution guidelines when submitting code

Acknowledgments

• Mistral AI for providing the language model API
• The IRC community for continued support
• All contributors to the project