How It Works

Multi-chain divergence scanner with volume proxy analysis and Wyckoff phase classification

Wyckoff Phases

Phase	SM Flow	Price	Signal
ACCUMULATION	Buying	Falling	Bullish divergence — SM buying the dip
DISTRIBUTION	Selling	Rising	Bearish divergence — SM exiting into strength
MARKUP	Buying	Rising	Trend confirmed — momentum aligned
MARKDOWN	Selling	Falling	Capitulation — both SM and price declining

Features

Dashboard

Real-time divergence heat map, signal feed, and metric cards. Tokens colored by Wyckoff phase, sized by market cap.

Pre-Breakout Radar

High volume-activity tokens showing divergent signals. Catch institutional positioning before the crowd.

Signal Performance

Track historical signal outcomes with win rates, return distributions, and timeline visualization.

Cross-Chain Flows

Chain momentum scores, sector rotation, and volume flow aggregation across 8 blockchains.

Token Deep Dive

Click any token in the dashboard table for full intelligence: flow by wallet type, buyers/sellers, Nansen Score, and wallet profiles.

Alpha Score

Each token gets an Alpha Score (0-100) based on four weighted factors: flow magnitude (40%), price movement (25%), wallet diversity (20%), and holdings conviction (15%). Scores use log-scaled normalization to prevent large-cap tokens from dominating.

Data Pipeline

1.MCP Discovery—Token search across 8 chains via Nansen MCP (0 credits)

2.CLI Enrichment—Real SM data via token screener + smart-money netflow (~12 credits/cycle)

3.Price History—SQLite-tracked price snapshots for real change computation

4.Volume Proxy—Vol/MCap ratio + relative volume → activity classification

5.Divergence Engine—Price-volume divergence → Wyckoff phase + confidence

6.Signal History—Outcome tracking + backtesting against price changes

Hybrid Pipeline: CLI + Volume Proxy

Top tokens on Ethereum and BNB Chain are enriched with real smart money data from the Nansen CLI (research token screener + research smart-money netflow). Remaining tokens use volume proxy analysis. Tokens with CLI data are marked with a greenCLI badge; volume-proxy tokens show a gray VP badge.

CLI Command	Data	Credits
research token screener	Real price, market cap, market netflow	1/page
research smart-money netflow	Real SM net flow per token (24h/7d)	5/page

Enrichment runs every 30 minutes, consuming ~12 credits per cycle (~288/day). Configurable viaCLI_ENRICH_CHAINS and CLI_ENRICH_MINUTES env vars.

Volume Proxy Methodology

Instead of requiring premium smart money wallet data, the scanner derives institutional activity signals from publicly available volume and price data:

Metric	Formula	What It Detects
Vol/MCap Ratio	volume_24h / market_cap	Unusual trading activity relative to token size
Relative Volume	current_vol / avg_vol_72h	Volume spikes vs recent baseline
Price-Vol Divergence	high_volume + falling_price	Accumulation (buying into weakness)
Whale Count Est.	volume / avg_institutional_trade	Estimated number of large transactions

Try It Yourself

Install the Nansen CLI and run these commands to see the raw data this scanner uses:

# Install Nansen CLI

npm i -g nansen-cli && nansen login

# Token screener — top tokens by volume (1 credit)

nansen research token screener --chain ethereum --timeframe 24h

# Smart money net flow — where SM wallets are moving (5 credits)

nansen research smart-money netflow --chain ethereum

# SM dex trades — individual whale trades (5 credits)

nansen research smart-money dex-trades --chain bnb

# Token deep dive — flow intelligence + Nansen Score (5 credits)

nansen research token flow-intelligence --chain ethereum --token 0x...

# Wallet profiler — who is this wallet? (1 credit)

nansen research profiler labels --address 0x... --chain ethereum

All 9 Nansen CLI endpoints are integrated. The scanner runs token screener +smart-money netflow every 30 minutes for ETH/BNB enrichment, plusflow-intelligence, who-bought-sold, andtoken indicators for top token deep dives.

Coverage

Scans 8 blockchains (Ethereum, BNB, Solana, Base, Arbitrum, Polygon, Optimism, Avalanche) via Nansen MCP discovery at zero credit cost. Auto-refreshed every 5 minutes. Signal outcomes tracked in SQLite and validated against subsequent price movements for backtesting.