🐕 AI Guard Dog Intern

Hackathon Project: Multi-agent AI system for DAO treasury protection on Monad blockchain

---

🎯 What is This?

AI Guard Dog Intern is a multi-agent AI system that automatically analyzes DAO proposals and protects treasury funds from malicious or poorly-written proposals. It runs on the Monad blockchain for high-speed, low-cost governance.

The Problem

DAOs face constant threats:

• 💀 Malicious proposals draining treasury
• 📝 Low-quality proposals wasting voter time
• ⏰ Voter fatigue leading to poor decisions
• 🚨 Scams disguised as legitimate proposals

The Solution

Three AI agents work together to:

1. Analyze every proposal for red flags
2. Score risk levels automatically
3. Route proposals appropriately:
   • ✅ Auto-approve low-risk proposals
   • 👀 Flag medium-risk for human review
   • 🚫 Auto-reject high-risk scams

---

🏗️ Architecture

┌─────────────────────────────────────────────────────────────┐
│                     USER SUBMITS PROPOSAL                    │
└─────────────────────────┬───────────────────────────────────┘
                          │
                          ▼
┌─────────────────────────────────────────────────────────────┐
│                    monad-dao/                               │
│  ┌─────────────┐  ┌─────────────┐  ┌─────────────────────┐  │
│  │ DAOToken    │  │ Treasury    │  │ ProposalManager     │  │
│  │ (Governance)│  │ (Funds)     │  │ (Lifecycle)         │  │
│  └─────────────┘  └─────────────┘  └─────────────────────┘  │
│  ┌─────────────┐  ┌─────────────┐  ┌─────────────────────┐  │
│  │ VotingEngine│  │ Timelock    │  │ AIAgentRegistry     │  │
│  │ (Tallying)  │  │ (Delay)     │  │ (AI Whitelist)      │  │
│  └─────────────┘  └─────────────┘  └─────────────────────┘  │
└─────────────────────────┬───────────────────────────────────┘
                          │ ProposalCreated Event
                          ▼
┌─────────────────────────────────────────────────────────────┐
│                   AI GUARD DOG (GuardDao/)                  │
│  ┌─────────────────────────────────────────────────────┐    │
│  │ Agent 1: Reputation Checker                         │    │
│  │   - Wallet age, history, ENS verification           │    │
│  └─────────────────────────────────────────────────────┘    │
│  ┌─────────────────────────────────────────────────────┐    │
│  │ Agent 2: NLP Analyzer                               │    │
│  │   - Red flag detection, scam patterns               │    │
│  └─────────────────────────────────────────────────────┘    │
│  ┌─────────────────────────────────────────────────────┐    │
│  │ Agent 3: Mediator                                   │    │
│  │   - Combines scores, makes routing decision         │    │
│  └─────────────────────────────────────────────────────┘    │
└─────────────────────────┬───────────────────────────────────┘
                          │ Risk Score
                          ▼
┌─────────────────────────────────────────────────────────────┐
│                    ROUTING DECISION                         │
│  Risk < 20:   ✅ Auto-Approve (VotingAgent votes FOR)       │
│  Risk 20-79:  👀 Human Review (Flag for DAO members)        │
│  Risk ≥ 80:   🚫 Auto-Reject (VotingAgent votes AGAINST)    │
└─────────────────────────────────────────────────────────────┘

---

📁 Project Structure

GuardDao/
├── contracts/                    # AI Guard Dog contracts
│   ├── VotingAgent.sol          # AI agent that casts votes
│   ├── AuditLogger.sol          # Logs all AI decisions
│   ├── DAOGovernor.sol          # Governor interface
│   └── MockToken.sol            # Test token
│
├── monad-dao/                   # Core DAO implementation
│   ├── contracts/
│   │   ├── core/
│   │   │   ├── DAOCore.sol      # Main coordinator
│   │   │   ├── DAOToken.sol     # Governance token
│   │   │   └── Treasury.sol     # Fund management
│   │   ├── governance/
│   │   │   ├── ProposalManager.sol  # Proposal lifecycle
│   │   │   ├── VotingEngine.sol     # Vote tallying
│   │   │   └── Timelock.sol         # Execution delay
│   │   ├── membership/
│   │   │   └── MemberRegistry.sol   # Member profiles
│   │   ├── integration/
│   │   │   └── AIAgentRegistry.sol  # AI agent whitelist
│   │   └── interfaces/              # All interfaces
│   │
│   └── scripts/
│       ├── deploy.js            # Deploy all contracts
│       ├── setup-dao.js         # Configure DAO
│       ├── fund-treasury.js     # Add funds
│       ├── register-ai-agent.js # Connect AI
│       └── create-test-proposal.js  # Demo proposals
│
├── PRD_AI_Guard_Dog_Intern.md   # Product requirements
├── Proposal_examples.md         # Sample proposals
└── README.md                    # This file

---

🚀 Quick Start

Prerequisites

• Node.js v18+
• npm or yarn

1. Install Dependencies

# Main project
cd GuardDao
npm install

# DAO contracts
cd monad-dao
npm install

2. Compile Contracts

cd monad-dao
npx hardhat compile

3. Start Local Blockchain

# Terminal 1 - Keep running
npx hardhat node

4. Deploy & Test

# Terminal 2
cd monad-dao
npx hardhat run scripts/deploy.js --network localhost
npx hardhat run scripts/setup-dao.js --network localhost
npx hardhat run scripts/fund-treasury.js --network localhost
npx hardhat run scripts/create-test-proposal.js --network localhost

---

📜 Smart Contracts

Core DAO Contracts (monad-dao/)

Contract	Address (localhost)	Purpose
DAOToken	deployment.json	Governance token with voting power
Treasury	deployment.json	Holds and manages DAO funds
ProposalManager	deployment.json	Creates and tracks proposals
VotingEngine	deployment.json	Handles vote casting and tallying
Timelock	deployment.json	Delays execution for security
MemberRegistry	deployment.json	Tracks member profiles
DAOCore	deployment.json	Coordinates all contracts
AIAgentRegistry	deployment.json	Whitelists AI agents

AI Guard Dog Contracts (contracts/)

Contract	Purpose
VotingAgent.sol	AI agent that analyzes and votes on proposals
AuditLogger.sol	Immutable log of all AI decisions

---

🔒 Security Features

• ⏰ Timelock: 2-day delay before execution
• 💰 Treasury Limits: 20% daily spending cap
• 🤖 AI Monitoring: Automatic red flag detection
• 📝 Audit Trail: All AI decisions logged on-chain

---

🧪 Test Proposals

The demo creates 3 proposals with different risk levels:

Proposal	Amount	Risk Level	Expected AI Action
Q1 Marketing	10 MON	Low (~15)	✅ Auto-Approve
Emergency Audit	50 MON	Medium (~45)	👀 Human Review
Partnership Deal	500 MON	High (~90)	🚫 Auto-Reject

---

🔗 Networks

Local Development

// hardhat.config.js
localhost: {
  url: "http://127.0.0.1:8545"
}

Monad Testnet (Coming Soon)

monadTestnet: {
  url: process.env.MONAD_TESTNET_RPC,
  accounts: [process.env.PRIVATE_KEY]
}

---

📊 Risk Scoring

The AI analyzes proposals using:

Red Flag Phrases

Phrase	Severity	Score Impact
"guaranteed returns"	Critical	+30
"trust us"	Warning	+10
"urgent/ASAP"	Warning	+5
"anonymous team"	Critical	+25
"confidential details"	Critical	+20

Quality Checks

• Budget breakdown completeness
• Team verification
• External link validity
• Historical pattern matching

---

🛣️ Roadmap

• Core DAO contracts
• Deployment scripts
• Test proposals
• AI risk scoring service
• Frontend dashboard
• Monad testnet deployment
• Production audit

---

🤝 Contributing

1. Fork the repository
2. Create feature branch (git checkout -b feature/amazing)
3. Commit changes (git commit -m 'Add amazing feature')
4. Push to branch (git push origin feature/amazing)
5. Open Pull Request

---

📄 License

MIT License - see LICENSE for details.

---

🙏 Acknowledgments

• Monad Team for the high-performance blockchain
• OpenZeppelin for security patterns
• Hackathon organizers