This is a Machine Learning Streamlit-powered dashboard that scans Ethereum blocks to detect and explain suspicious on-chain activity. This tool helps users identify patterns commonly associated with bots, including:
• High Gas Transactions
• Sandwich Attacks
• Anomalous Transactions
• MEV Bot Clusters
This project is useful for traders, developers, and analysts who want a clearer view of hidden activity on the Ethereum network.
• Beginner-friendly explanations for each suspicious behavior
• Real-time data fetched from the Ethereum blockchain
• Custom thresholds for high gas price detection
• Live counters for high gas transactions, sandwich attacks, and anomalies
• Secure API integration using Streamlit secrets
• Performance optimized to fetch and analyze up to 100 blocks per scan within 60 seconds
The dashboard fetches a set number of recent Ethereum blocks using the Etherscan API. It analyzes transactions to:
• Detect high gas usage based on a defined threshold which is 3 time the current median of gas prices
• Identify sandwich attacks by comparing gas prices and execution order
• Flag transactions with low transfer value but unusually high gas costs
• Group similar suspicious transactions into MEV bot clusters
• Python
• Streamlit
• Pandas
• Requests
• Etherscan API