Developer Launches Open-Source Tool to Detect Crypto Pump-and-Dump Schemes

Developer aleks-ent has released an open-source tool designed to identify cryptocurrency pump-and-dump regimes. The project, titled 'pump-dump-crypto-screener,' provides a transparent method for detecting artificial price inflation in volatile markets.

Built as a TypeScript pnpm monorepo, the tool analyzes 1-minute and 5-minute market statistics and kline archives. According to the project's GitHub repository, the screener relies exclusively on 'candle-only' data sourced from Binance and Bybit to trigger its detections. To facilitate immediate action, the developer integrated a Telegram bot, @pumpdumpscreenerautobot, which provides real-time alerts to users when suspicious patterns emerge. The technical stack requires Node.js 20+ and pnpm 9+.

The Shift to Price Action

Pump-and-dump schemes are a persistent issue in low-liquidity cryptocurrency markets, where coordinated groups artificially inflate a token's price before rapidly selling off their holdings. Traditionally, most detection tools have relied on social media sentiment analysis or order book imbalances to spot these events. However, social data is often noisy or lagged, and order book data can be manipulated by spoofing.

By focusing strictly on kline archives—the historical price candles—this tool ignores external noise and focuses on the actual market result. This approach treats the price action itself as the primary signal of manipulation, removing the dependency on unreliable external indicators.

Implications for Traders

Open-sourcing a candle-only detector provides a significant advantage for retail traders who typically rely on proprietary 'black box' signals. Because the logic is public, users can verify exactly how a pump is identified rather than trusting a closed-source provider. This transparency reduces the reliance on external social data, which is frequently used by the orchestrators of pump-and-dump schemes to lure in unsuspecting buyers.

Future Development

While the core detection engine is functional, the project is currently in an expansion phase. According to a post on Hacker News, the creator is actively seeking machine learning contributors to enhance the tool's detection capabilities. The goal is to move beyond static statistics and implement more sophisticated ML models to improve the accuracy of the alerts and reduce false positives in highly volatile markets, ensuring that traders can better distinguish between organic growth and coordinated manipulation.