Address Clustering Tool

WHAT IT DOES

This tool identifies address clusters—groups of wallet addresses that are likely controlled by the same entity—by analyzing transaction patterns and interaction histories. It reads transaction history for the connected wallet via on-chain calls, then applies clustering algorithms to group related addresses. The tool supports three clustering methods: Shared Contract (addresses that interact with the same contracts), Transaction Pattern (addresses with similar transaction behaviors), and Graph Connectivity (addresses connected through transaction flows). You can control analysis depth and minimum transaction thresholds to refine results. The output includes total addresses analyzed, total connections, clusters found, seed cluster ID, and lists of related addresses within each cluster.

HOW TO USE

Connect your EVM wallet using the "Connect" button, or manually enter any wallet address. Select the clustering algorithm (Shared Contract, Transaction Pattern, or Graph Connectivity). Set the analysis depth (1–5 hops) and minimum transactions per cluster (1–10). Select your network and click "Find address clusters" to run the analysis. The tool will display the source address, network, algorithm used, depth, min transactions, total addresses, total connections, clusters found, seed cluster ID, related addresses in the same cluster (with labels), and a summary of all clusters with member counts. The seed cluster (containing your source address) is marked with a star.

TECHNICAL MECHANISM

The tool uses graph-based clustering algorithms to group related addresses. First, it fetches transaction history for the source address via on-chain calls (simulated in this demo, but would use ethers.js provider calls with getLogs and getTransactionReceipt in production). The data is modeled as a directed graph where addresses are nodes and transactions are edges. Three clustering approaches are implemented: (1) Shared Contract—addresses that interact with the same contract addresses are grouped; (2) Transaction Pattern—addresses with similar transaction frequencies, value ranges, and timing patterns are grouped using behavioral similarity; (3) Graph Connectivity—addresses connected through transaction flows are grouped using connected components analysis. The depth parameter controls how far from the source address the traversal extends. The minimum transactions parameter filters out clusters with fewer than the specified number of transactions. The seed cluster is identified as the cluster containing the source address.

WHAT IT CANNOT SEE

The tool cannot view transactions on non-EVM chains (e.g., Solana, Bitcoin) or protocols that are not integrated with the tool's data sources. It cannot show real-time pending transactions or mempool activity—only confirmed on-chain transaction history is available. It cannot access wallet data without explicit user connection or manual address entry. It cannot detect off-chain holdings or centralized exchange balances. Clustering is limited to available historical data; older transactions may not be fully retrievable due to node limitations or RPC provider retention policies. The tool does not guarantee complete address clustering—some addresses may be missed due to data gaps or incomplete transaction histories. The tool's data sources may have rate limits or require API keys for production use. It does not provide financial advice or guarantee the accuracy of aggregated data.

PLEASE NOTE

This tool is for informational and visualization purposes only and is not financial advice. Address clustering is based on historical data and may not reflect current address relationships. Clustering algorithms are heuristic and may produce false positives or miss related addresses. Always verify cluster relationships directly on-chain using a block explorer. The tool's data sources are subject to rate limits and availability. This tool only works with EVM-compatible chains. Use at your own risk.