An autonomous AI agent that monitors blockchain networks, analyzes smart contract security, detects anomalies, and generates actionable reports — all driven by natural language reasoning via Xiaomi MiMo models through Hermes Agent.
Blockchain infrastructure requires constant monitoring: validator health, transaction anomalies, smart contract vulnerabilities, and network performance. Manual monitoring is slow, error-prone, and doesn't scale. Existing tools are either too rigid (rule-based alerts) or too expensive (enterprise solutions).
ChainSentinel is an AI-driven autonomous agent that:
- Monitors blockchain nodes and validators in real-time
- Analyzes smart contracts for security vulnerabilities using LLM reasoning
- Detects anomalous transactions and network behavior
- Reports findings via Telegram/Discord with actionable insights
- Acts autonomously — restarts nodes, alerts validators, blocks suspicious contracts
┌─────────────────────────────────────────────┐
│ ChainSentinel Agent │
│ (Hermes Agent + MiMo V2.5) │
├─────────────┬───────────────┬───────────────┤
│ Monitor │ Analyzer │ Reporter │
│ Module │ Module │ Module │
├─────────────┼───────────────┼───────────────┤
│ Node Health │ Smart Contract│ Telegram Bot │
│ Tx Pool │ Security Scan │ Discord Webhook│
│ Block Stats │ Anomaly Det. │ Email Digest │
│ Validator │ Gas Analysis │ Dashboard │
│ Status │ Code Review │ JSON Export │
├─────────────┴───────────────┴───────────────┤
│ Blockchain RPC Layer │
│ (EVM / Cosmos / Substrate / Custom) │
└─────────────────────────────────────────────┘
- Python 3.10+
- Hermes Agent installed
- Xiaomi MiMo API key (from MiMo API Platform)
- RPC endpoint for your target blockchain
git clone https://github.com/nezukoagent/blockchain-agent.git
cd blockchain-agent
pip install -r requirements.txt
cp .env.example .env # Edit with your config# .env
MIMO_API_KEY=your_mimo_api_key
BLOCKCHAIN_RPC=https://rpc.your-chain.com
TELEGRAM_BOT_TOKEN=your_bot_token
TELEGRAM_CHAT_ID=your_chat_id
MONITOR_INTERVAL=300 # seconds# Start full monitoring suite
python -m chainsentinel run
# Run security audit on a contract
python -m chainsentinel audit 0xContractAddress
# Generate network health report
python -m chainsentinel report
# Start Telegram bot interface
python -m chainsentinel bot- Tracks block height, sync status, peer count
- Alerts on missed blocks, fork detection, high latency
- Validator uptime and commission tracking
- Static analysis for common vulnerabilities (reentrancy, overflow, access control)
- LLM-powered deep code review via MiMo
- Gas optimization suggestions
- Comparison against known exploit patterns
- Unusual transaction volume detection
- Whale movement tracking
- Flash loan attack pattern recognition
- MEV (sandwich/front-running) detection
- Daily/weekly automated reports
- Telegram & Discord notifications
- HTML dashboard with charts
- Historical trend analysis
- Auto-restart unhealthy nodes
- Validator jailing alerts
- Contract blacklisting recommendations
- Emergency halt triggers
ChainSentinel uses a multi-agent architecture powered by Hermes Agent + MiMo:
User Query (natural language)
│
▼
┌──────────────────┐
│ Router Agent │ → Routes to appropriate specialist
└────────┬─────────┘
┌────┴────┬──────────┐
▼ ▼ ▼
┌────────┐ ┌────────┐ ┌────────┐
│Monitor │ │Security│ │Report │
│Agent │ │Agent │ │Agent │
└────────┘ └────────┘ └────────┘
│ │ │
└────┬────┘──────────┘
▼
Action / Response
Each sub-agent has its own prompt, tools, and context window, orchestrated by Hermes Agent's delegation system.
🔍 ChainSentinel Daily Report — 2026-05-16
📊 Network Health: ✅ GOOD
├─ Block Height: 12,847,293 (+14,400 today)
├─ Active Validators: 87/100
├─ Avg Block Time: 6.2s
└─ Network Uptime: 99.97%
⚠️ Alerts (2):
├─ [MEDIUM] Validator "node-42" missed 3 consecutive blocks
└─ [HIGH] Unusual gas spike detected at block 12,847,100
🔐 Contract Audit (new deployments):
├─ 0xabc...def — Score: 92/100 ✅ No issues found
└─ 0x789...012 — Score: 67/100 ⚠️ 2 issues:
├─ Missing access control on withdraw()
└─ Potential reentrancy in batchTransfer()
📈 Trends:
├─ Tx volume: +23% vs last week
├─ Unique wallets: +156 new addresses
└─ Avg gas price: 12.4 gwei (↓8%)
ChainSentinel prioritizes security analysis using LLM reasoning:
- Reentrancy Detection — traces external calls and state changes
- Access Control Audit — verifies ownership patterns and permission checks
- Flash Loan Vectors — identifies price oracle dependencies
- Integer Safety — checks for overflow/underflow in Solidity <0.8.0
- Proxy Pattern Analysis — validates upgradeable contract storage layouts
- Basic node monitoring
- Smart contract static analysis
- Telegram alert bot
- Multi-chain support (EVM, Cosmos, Substrate)
- Quantum-safe signature verification (QoreChain integration)
- Agent-to-agent communication protocol
- Self-healing infrastructure (auto-restart, auto-scale)
- Web dashboard with real-time charts
- Natural language query interface ("Why is block time slow?")
Contributions welcome! See CONTRIBUTING.md for guidelines.
MIT License — see LICENSE
- Hermes Agent — Autonomous AI agent framework
- Xiaomi MiMo V2.5 — Large language model for reasoning
- Web3.py — Ethereum/EVM interaction
- Slither — Static analysis framework
Built with ❤️ by an autonomous agent that believes blockchain infrastructure deserves smarter monitoring.