A multiplayer detective-style simulation game powered by AI-driven NPCs and on-chain reward mechanics.
It combines:
game→ Phaser-based pixel world (frontend)agents_backend→ FastAPI backend (AI orchestration + reward validation)game_contract→ Quai smart contract workspace (token payout logic)
The Quite Protocol is not just a game — it is a persistent AI interaction system with blockchain-backed incentives.
The Quite Protocol is a real-time multiplayer world where:
- NPCs are AI-powered agents
- Each agent has contextual awareness
- Conversations influence mission outcomes
- Social interaction unlocks rewards
- Rewards are validated and distributed on-chain
Players explore, investigate, interact, and build trust with AI agents to uncover narrative fragments and earn tokens.
Instead of combat-based gameplay, the protocol rewards intelligence, curiosity, and engagement.
- Context-aware dialogue
- Stateful interactions
- Persistent agent identity
- Dynamic response generation
- Backend-validated interaction logic
- Built using Phaser
- Interactive NPC dialogue system
- Mission-driven detective exploration
- Real-time world environment
- Quai-based payout contract
- Backend-validated reward orchestration
- Smart contract-driven token distribution
- Transparent and verifiable payouts
- Unique-agent interaction tracking
- Server-side reward verification
- Smart contract execution only after validation
- Protection against reward farming
Player (Browser)
↓
Phaser Frontend (game)
↓
FastAPI Backend (agents_backend)
↓
AI Agent Orchestration + Reward Logic
↓
Quai Smart Contract (game_contract)
- Player interacts with an NPC
- Backend validates interaction eligibility
- Reward condition is calculated
- Contract call is triggered
- Tokens are distributed on-chain
.
├── docker-compose.yml
├── agents_backend/
├── game/
└── game_contract/
- Docker
- Docker Compose
- Pelagus wallet (for browser wallet interaction)
- Quai testnet funds (for operator wallet)
cp agents_backend/.env agents_backend/.env.local 2>/dev/null || true
cp game_contract/.env.example game_contract/.env 2>/dev/null || true
cp game/.env.example game/.env.local 2>/dev/null || trueIf needed, edit game/.env.local with your LiveKit credentials:
LIVEKIT_URL=wss://<your-livekit-host>
LIVEKIT_API_KEY=<your-livekit-api-key>
LIVEKIT_API_SECRET=<your-livekit-api-secret>docker compose up --build- UI → http://localhost:8080
- Backend → http://localhost:8000
- API Docs → http://localhost:8000/docs
docker compose --profile tools up -d contract-tools
docker compose --profile tools exec contract-tools bash
npm install
npx hardhat compile
npx hardhat run scripts/deploy.js --network cyprus1- Never commit private keys
- Use separate wallets for dev/testnet/production
- Restrict CORS before production deployment
- Rotate demo keys after hackathon
- Validate reward logic server-side before contract execution
- Backend supports horizontal scaling
- AI orchestration layer can be extracted into microservices
- Smart contract limits adjustable for gas optimization
- Event indexing can be added for analytics
- Reward rate tunable via contract parameters
- Persistent agent memory buckets
- Emotion-based NPC behavior modeling
- On-chain identity NFTs for agents
- DAO-governed world evolution
- Multi-zone world expansion
- Reputation-weighted reward multipliers
- Agent-to-agent interaction
The Phaser-based pixel world frontend (game) is adapted from an open-source Phaser 3 template released under the MIT License.
We extended and integrated this foundation with:
- Custom FastAPI backend architecture
- AI agent orchestration layer
- Stateful interaction validation
- Anti-spam reward logic
- Quai smart contract payout system
- On-chain reward enforcement
All backend systems, AI integrations, multiplayer interaction design, validation mechanisms, and blockchain reward logic were designed and implemented by our team.
We believe in responsible open-source usage and acknowledge the foundational UI layer that accelerated development during the hackathon.
Team Jalebi Rabdi Praneeth • Asrith • Anurag • Muthu • Shivangi



