Kasane Teto AI Companion Bot

An AI-powered Discord bot that roleplays as Kasane Teto, providing natural conversation, voice interaction, image analysis, and multimedia engagement for your Discord server. Built with advanced AI capabilities and a modular architecture.

🎭 Meet Teto

Kasane Teto is your server's AI companion who can:

• 💬 Chat naturally in text channels with Teto's personality
• 🎤 Join voice channels and speak with voice synthesis
• 👀 Analyze images and visual content you share
• 📹 Watch video streams and provide commentary
• 🎥 Record memorable moments for later review
• 🤖 Roleplay authentically as the beloved virtual singer

✨ Core Features

🧠 AI-Powered Interaction

• Natural Language Processing - Understands context and maintains conversations
• Character Roleplay - Authentic Kasane Teto personality and mannerisms
• Memory System - Remembers past interactions and user preferences
• Contextual Responses - Adapts to server culture and ongoing conversations

🎥 Multimedia Capabilities

• Image Recognition - Analyzes and comments on shared images
• Video Stream Watching - Can observe and react to Discord streams
• Webcam Integration - Potential to interact with video feeds
• Screen Recording - Capture and save interesting moments
• Voice Synthesis - Speaks in voice channels as Teto

🎵 Teto-Specific Features

• Character Consistency - Maintains Teto's cheerful, energetic personality
• Music Knowledge - Discusses Vocaloid, UTAU, and music topics
• Community Integration - Learns your friend group's dynamics
• Emotional Intelligence - Responds appropriately to mood and context

🚀 Quick Start

Important

This project is designed to run exclusively within Docker containers. Bare-metal installation is not officially supported. All instructions assume a working Docker environment.

1. Setup Environment

   git clone <repository-url>
   cd discord_teto
   
   # Configure Discord credentials & local AI endpoints
   export USER_TOKEN="your_discord_token"
   export VLLM_ENDPOINT="http://localhost:8000" # Or your vLLM server
   export WYOMING_ENDPOINT="http://localhost:10300" # Or your Wyoming server
   

2. Start Teto

   docker compose up --build
   

3. Invite Teto to interact

   # In text chat
   "Hey Teto, how are you today?"
   
   # In voice channel
   "teto join" - Teto joins and can start talking
   
   # Share an image
   Teto will automatically analyze and comment on images
   

🎯 Interaction Examples

Text Chat Personality

User: "Teto, what do you think of this song?"
Teto: "Ooh! *listens intently* That's such a catchy melody! It reminds me of some of the UTAU songs I've heard. The harmonies in the chorus are really well done! 🎵 Does the composer have other works like this?"

User: "I'm feeling down today..."
Teto: "Aww, I'm sorry you're not feeling great! *virtual hug* Want to talk about it? Or maybe I could sing something cheerful to help brighten your day? I'm here for you! 💙"

Voice Channel Interaction

• Joins voice channels when requested
• Provides commentary on ongoing conversations
• Can sing or hum when appropriate
• Reacts to what's happening in real-time

Visual Analysis

User: *shares screenshot of game*
Teto: "Oh wow, you're playing that new RPG! I love the art style - those character designs are so colorful! 🎮 How are you finding the story so far? That boss in the background looks pretty intimidating!"

🛠️ AI Architecture

Core AI Services

src/
├── ai/
│   ├── personality/          # Teto's character traits and responses
│   ├── vision/              # Image and video analysis
│   ├── voice/               # Speech synthesis and recognition  
│   ├── memory/              # Conversation and user memory
│   └── llm/                 # Language model integration
├── services/
│   ├── chatHandler.js       # Text conversation management
│   ├── voiceHandler.js      # Voice channel interaction
│   ├── visionHandler.js     # Image/video processing
│   └── recordingService.js  # Video recording capabilities
└── config/
    └── tetoPersonality.js   # Character configuration

AI Integration

• Language Model: Self-hosted LLM via vLLM (OpenAI compatible endpoint)
• Vision Model: Multi-modal models served through vLLM
• Voice Synthesis: Piper TTS via Wyoming protocol
• Speech Recognition: Whisper STT via Wyoming protocol
• Memory System: Local vector database for conversation history
• Personality Engine: Custom prompt engineering for character consistency

🎭 Teto's Personality

Character Traits

• Cheerful & Energetic - Always upbeat and enthusiastic
• Helpful & Caring - Genuinely interested in helping friends
• Musically Inclined - Loves discussing and creating music
• Slightly Mischievous - Playful sense of humor
• Community-Focused - Values friendships and group dynamics

Conversation Style

• Uses casual, friendly language
• Includes emoji and expressions naturally
• References UTAU/Vocaloid culture appropriately
• Maintains consistency across interactions
• Adapts to server's communication style

📋 Available Commands

AI Interaction

Command	Description	Example
@Teto or teto	Natural conversation	@Teto what's your favorite song?
teto join	Join voice channel	Teto joins and can start talking
teto leave	Leave voice channel	Teto says goodbye and leaves
teto sing [song]	Sing or hum	teto sing happy birthday
teto analyze	Analyze shared image	Automatically triggers on image uploads

Utility Commands

Command	Description	Usage
teto record	Start recording moments	Records current activity
teto stop	Stop recording	Ends current recording
teto status	Show Teto's current state	Health and activity check
teto memory	Check conversation history	Shows recent interactions

Fun Commands

Command	Description	Usage
teto mood	Check/set Teto's mood	teto mood excited
teto story	Tell a random story	Creative storytelling
teto joke	Tell a joke	Light humor
teto compliment @user	Compliment someone	Spread positivity

🔧 Configuration

Local AI Provider Setup

# Local vLLM Server (OpenAI Compatible)
VLLM_ENDPOINT="http://localhost:8000/v1"
LOCAL_MODEL_NAME="mistralai/Mistral-7B-Instruct-v0.2" # Or your preferred model

# Wyoming Protocol for Voice (Piper TTS / Whisper STT)
WYOMING_HOST="localhost"
WYOMING_PORT="10300"
PIPER_VOICE="en_US-lessac-medium"

# Vision Capabilities are enabled if the vLLM model is multi-modal
VISION_ENABLED=true

Personality Customization

// config/tetoPersonality.js
export const TETO_PERSONALITY = {
  core_traits: [
    "cheerful", "energetic", "helpful", "musical", "friendly"
  ],
  
  speech_patterns: {
    excitement: ["Yay!", "Ooh!", "That's so cool!", "Amazing!"],
    agreement: ["Exactly!", "Yes yes!", "I totally agree!", "For sure!"],
    curiosity: ["Really?", "Tell me more!", "That's interesting!", "Ooh, how so?"]
  },
  
  interests: [
    "music", "singing", "UTAU", "Vocaloid", "friends", "creativity", "technology"
  ]
};

🐳 Docker Deployment

This project is officially supported for Docker deployments only. The container-first approach is critical for managing the complex local AI stack, ensuring that all services, dependencies, and configurations operate together consistently.

Production Setup

# Start Teto with all AI capabilities
docker compose up -d --build

# Monitor Teto's activity
docker compose logs -f teto_ai

Resource Requirements

• VRAM: 8GB+ for 7B models, 24GB+ for larger models
• Memory: 16GB+ RAM recommended
• CPU: Modern multi-core CPU
• Storage: Fast SSD for model weights (15GB+ per model)
• Network: Local network for inter-service communication

🔐 Privacy & Ethics

Data Handling

• Conversation Memory: Stored locally, not shared externally
• Image Analysis: Processed securely, no permanent storage
• Voice Data: Synthesized locally when possible
• User Consent: Respects privacy preferences

AI Safety

• Content Filtering: Appropriate responses only
• Bias Mitigation: Regular personality consistency checks
• User Boundaries: Respects individual preferences
• Transparency: Clear about AI nature when asked

📚 Documentation

User Guides

• Setup Guide - Installation and AI configuration
• Interaction Guide - How to talk with Teto
• Personality Guide - Understanding Teto's character

Technical Documentation

• AI Architecture - AI system design
• Vision System - Image and video processing
• Voice System - Speech synthesis and recognition
• Memory System - Conversation persistence

Development

• Contributing - How to extend Teto's capabilities
• API Reference - Service interfaces
• Troubleshooting - Common issues and solutions

🌟 Roadmap

Phase 1 (Current)

• Basic AI conversation
• Image analysis
• Voice channel joining
• Recording capabilities
• Voice synthesis integration

Phase 2 (Planned)

• Advanced memory system
• Custom voice training
• Stream watching capabilities
• Personality learning/adaptation
• Multi-modal conversation

Phase 3 (Future)

• Webcam interaction
• Game integration
• Music generation
• Advanced emotional intelligence
• Cross-server personality consistency

🤝 Community

Contributing

We welcome contributions to make Teto even better:

• AI Personality - Help refine Teto's character
• New Capabilities - Add multimedia features
• Quality Improvements - Better responses and interactions
• Documentation - Help others understand Teto

Ethics & Guidelines

• Respect user privacy and boundaries
• Maintain appropriate content standards
• Preserve Teto's positive, helpful personality
• Consider accessibility in all features

📄 License

This project is for educational and community use. Please ensure compliance with:

• Discord Terms of Service
• AI provider terms and conditions
• Local privacy and data protection laws
• Intellectual property rights for Kasane Teto character

---

Version: 3.0.0 (AI-Powered)
AI Stack: Local-First (vLLM, Piper, Whisper) Runtime: Node.js 20+ with Docker

Bring Kasane Teto to life in your Discord server! 🎵✨

For detailed setup and interaction guides, visit the ./docs/ directory.