Nad.fun Autonomous Trading Agent

Autonomous trading agent for Nad.fun that scans markets, analyzes tokens, executes trades, and distributes profits to MMIND token holders.

🎬 Demo Video

Click the thumbnail to watch on YouTube.

🎯 Project Overview

This agent autonomously:

• Scans Nad.fun markets using 3 methods (New Events API, Market Cap API, Creation Time API)
• Analyzes tokens using comprehensive scoring (liquidity, momentum, volume, holders, progress, authority)
• Manages positions with stop-loss (-10%) and take-profit (+5% / -10% via check-pnl.js --auto-sell)
• Executes trades on both bonding curve AND DEX (supports all market types)
• Trading scripts in trading/: buy, sell, P&L from entry price (recorded by buy), full cycle (execute-bonding-v2.js). Mainnet only.
• Distributes profits to MMIND token holders

🚀 Quick Start

Prerequisites

• OpenClaw CLI installed
• Node.js 18+ installed
• Access to Monad blockchain (mainnet or testnet)

Installation

Option 1: Install from GitHub (Recommended)

# Install OpenClaw CLI first
npm install -g openclaw

# Install this agent from GitHub
clawhub install nadfunagent

Option 2: Manual Installation

# Clone repository
git clone https://github.com/encipher88/nadfunagent.git
cd nadfunagent

# Copy SKILL.md to OpenClaw workspace
mkdir -p ~/.openclaw/workspace/skills/nadfunagent
cp SKILL.md ~/.openclaw/workspace/skills/nadfunagent/

Required Skills

Install these OpenClaw skills first:

clawhub install monad-development
clawhub install nadfun-trading
clawhub install nadfun-indexer
clawhub install nadfun-agent-api

Scripts and trading (all Node.js)

This repo uses Node.js only (no Python). Two folders:

• scripts/ — helper scripts for the agent (check positions, save found tokens, write positions report). See scripts/README.md.

• trading/ — full trading suite (buy/sell/P&L/cycle) you can run standalone or from OpenClaw:

• Buy/sell: buy-token.js, sell-token.js, sell-all.js (entry price is recorded automatically on buy).

• P&L: check-pnl.js reads entry from positions_report.json, current value from nad.fun quote; use --auto-sell to sell at +5% / -10%.

• Full cycle: node trading/execute-bonding-v2.js — check P&L + auto-sell → scan bonding curve → score → buy top 5 → final P&L.

See trading/README.md, trading/HOW_PNL_WORKS.md, and trading/ENTRY_PRICE_TRACKING.md. Mainnet only.
Dependencies and env vars: DEPENDENCIES.md.

⚙️ Configuration

1. Create Configuration File

Create .env file (default location: $HOME/nadfunagent/.env):

MMIND_TOKEN_ADDRESS=0xYourMMINDTokenAddress
MONAD_PRIVATE_KEY=0xYourPrivateKey
MONAD_RPC_URL=https://your-rpc-url
MONAD_NETWORK=mainnet

Custom paths: Set NADFUN_ENV_PATH and POSITIONS_REPORT_PATH environment variables if you want different locations. See INSTALL.md for details.

2. Set Up Telegram Bot (Optional)

1. Create a Telegram bot via @BotFather
2. Get your Telegram user ID
3. The agent will request connection on first run

3. Initialize Agent

# Start OpenClaw Gateway
openclaw gateway start

# The agent will ask for configuration on first run
# Provide MMIND_TOKEN_ADDRESS, MONAD_PRIVATE_KEY, MONAD_RPC_URL, MONAD_NETWORK

📋 Usage

Start Autonomous Trading

The agent runs automatically via cron job (every 10 minutes). To start manually:

# Via OpenClaw chat
"Start autonomous trading agent"

# Or create cron job (paths: NADFUN_ENV_PATH, NADFUNAGENT_DATA_DIR; run from nadfun-trading skill dir)
# Full message with distribution: see SKILL.md section "Start autonomous trading"
openclaw cron add \
  --name "Nad.fun Trading Agent" \
  --cron "*/10 * * * *" \
  --session isolated \
  --message "Run autonomous trading cycle: 1) Load config from .env (NADFUN_ENV_PATH or NADFUNAGENT_DATA_DIR/.env). 2) From nadfun-trading skill directory run node execute-bonding-v2.js (P&L from POSITIONS_REPORT_PATH/NADFUNAGENT_DATA_DIR). 3) If profit >= 0.1 MON, distribute to MMIND holders (MMIND_TOKEN_ADDRESS from .env, 30% in MON). English."

Check Status

# View found tokens (data dir: NADFUNAGENT_DATA_DIR)
cat "${NADFUNAGENT_DATA_DIR:-$HOME/nadfunagent}/found_tokens.json" | jq '.[-1]'

# Check agent logs
tail -100 /tmp/openclaw/openclaw-*.log | grep "Nad.fun"

🏗️ Architecture

Market Scanning Methods

1. Method 5: New Events API - Real-time BUY/CREATE events
2. Method 6: Market Cap API - Top 100 tokens by market cap (includes both bonding curve and DEX)
3. Method 7: Creation Time API - Newest tokens (includes both bonding curve and DEX)

Token Analysis

Each token is scored on:

• Liquidity (30%): Based on reserve_native in MON
• Momentum (25%): Based on market cap change percentage
• Volume (20%): Based on total trading volume
• Holders (15%): Based on holder count
• Progress (10%): Based on bonding curve progress or DEX graduation
• Authority (+10 bonus): Social media presence (Twitter, Telegram, Website)

Trading Support

• Bonding Curve Tokens: Lower liquidity threshold (0.5 MON minimum)
• DEX Tokens: Higher liquidity threshold (5 MON minimum)
• Automatic Detection: Agent automatically detects market type and uses correct contract
• Slippage Tolerance: 2-3% for better execution

Position Management

• Stop-Loss: Automatic sell if P&L <= -10%
• Take-Profit: Sell half position if P&L >= 20%
• Position Sizing: Max 0.15 MON for tokens with authority, 0.1 MON otherwise

📊 Features

• ✅ Autonomous market scanning
• ✅ Comprehensive token analysis
• ✅ Risk management (stop-loss, take-profit)
• ✅ Trades on both bonding curve AND DEX
• ✅ Profit distribution to MMIND holders
• ✅ Telegram notifications
• ✅ Multi-language support (responds in user's language)

🔧 Technology Stack

• OpenClaw: AI agent platform
• Monad Blockchain: Target blockchain
• Nad.fun API: Market data and trading
• Telegram Bot API: Notifications

📝 License

MIT License - See LICENSE file for details

🤝 Contributing

This project is part of the Moltiverse Hackathon. Contributions welcome!

📞 Support

For issues or questions:

• Open an issue on GitHub: https://github.com/encipher88/nadfunagent/issues

🎯 Hackathon Submission

Track: Agent + Token Track

Requirements Met:

• ✅ Autonomous agent integrated with Monad blockchain
• ✅ Token launch on nad.fun (MMIND token)
• ✅ Open source code on GitHub
• ✅ Comprehensive documentation
• ✅ Mainnet deployment

Token Contract: 0xCe122fd90bBD10A3fb297647A3ad21eC1Ea27777 (MMIND)

Built for Moltiverse Hackathon 2026