Discord Bot

• Purpose: A small Discord bot with two primary purposes: a music player and a chat assistant.

Music Player

• /play: Play audio from YouTube (supports search terms or direct URLs). Adds to the guild queue and shows a playback view with controls.
• /join: Joins the voice channel you're in.
• /leave: Disconnects the bot from the voice channel.
• /queue: Shows the current queue and now-playing information.
• /skip: Skips the current song.
• /stop: Stops playback, clears the queue, and disconnects.
• /clear: Clears the queue.
• Button controls: A UI view with Skip, Pause/Resume, and Stop buttons appears alongside some music responses.

Chat Assistant

• The bot listens for mentions in text channels and forwards the message (plus recent channel context) to a local LLM HTTP API.
• The LLM endpoint is expected at http://localhost:11434/api/generate by default.
• The chat cog returns the model reply as a threaded reply in-channel.

Files of Interest

• main.py - Bot bootstrap, event handlers, and global commands such as /ping, /hello, /pfp, and /credits.
• cogs/music.py - Music cog: queue management, yt-dlp integration, and FFmpeg playback.
• cogs/chat.py - Chat cog: collects context, calls the local LLM API, and replies when mentioned.
• environment.yaml - Conda environment specification (replaces requirements.txt).

Setup & Run

1. Create the Conda environment from environment.yaml (from the repo root):

conda env create -f environment.yaml

2. Activate the environment (the environment name in this repo is discord-bot):

conda activate discord-bot

3. Create a .env file (do NOT commit this). Required variables:

• DISCORD_TOKEN - Your bot token
• DISCORD_GUILD - Guild ID where you want to register commands

Example .env (do not paste real tokens into source control):

DISCORD_TOKEN=YOUR_TOKEN_HERE
DISCORD_GUILD=YOUR_GUILD_ID

4. Install Ollama for the chat assistant feature.

5. Run the bot using the launch script:

./launch.sh

The launch script will:

• Kill any existing bot process (prevents duplicate responses)
• Start Ollama if not already running
• Launch the bot

Alternatively, run manually:

python main.py