Token Research Agent

🐋 Token Research Agent

An AI-powered crypto token security analyst. Give it any Ethereum or Solana token address and it will fetch on-chain data from multiple sources, run a structured risk analysis via Claude, and print a colour-coded report in the terminal.


What it does

1. Auto-detects chain — 0x… addresses → Ethereum, base58 → Solana
2. Fetches parallel data from:
   • DexScreener — price, liquidity, volume, pair age (ETH + SOL)
   • Etherscan — contract verification, holder count (ETH only)
   • Honeypot.is — honeypot flag, buy/sell/transfer tax (ETH only)
   • Solscan — token metadata, top-10 holder concentration (SOL only)
   • RugCheck — risk score and risk items (SOL only)
3. Runs an agentic Claude loop — Claude uses tool-use to request and reason over the data
4. Prints a rich terminal report with colour-coded risk tier

---

Stack

Layer	Technology
Language	Python 3.12
HTTP client	httpx (async)
AI	Anthropic SDK (anthropic)
Data validation	Pydantic v2
Terminal UI	rich
Container	Docker / Docker Compose

---

Setup

1. Clone and install

git clone https://github.com/your-username/token-research-agent.git
cd token-research-agent
python -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt

2. Configure environment

cp .env.example .env

Edit `.env` and fill in your keys:

Variable	Where to get it
ANTHROPIC_API_KEY	console.anthropic.com
ETHERSCAN_API_KEY	etherscan.io/apis

3. Run

# Ethereum token
python -m agent.main 0xA0b86991c6218b36c1d19D4a2e9Eb0cE3606eB48

# Solana token
python -m agent.main EPjFWdd5AufqSSqeM2qN1xzybapC8G4wEGGkZwyTDt1v

---