🛠️ Code Base

Explore Nemesia’s code on GitHub! The repository contains full source code, documentation, and instructions for setting up and contributing to the project.

Visit our GitHub Repository to get started.

🔬 Main Formulas

1. Rug Pull Prediction

🔍 AI models for transaction analysis

• Detects potential rug-pulls in DeFi projects by analyzing transaction patterns.

• Automatic token and smart contract checks for signs of fraud and manipulation.

Formula:

Suspicious Transaction Test:

function isSuspiciousTransaction(transactionVolume, avgVolume) {
  const threshold = 1.5;  // Example threshold for suspicion
  return (transactionVolume / avgVolume) > threshold;
}

Where:

transactionVolume — volume of the current transaction, avgVolume — average transaction volume, threshold — threshold for detecting suspicious activity.

AI Prediction for Rug Pull:

function predictRugPull(tokenData) {
  const features = [
    tokenData.liquidityChange,
    tokenData.unusualTransactionCount,
    tokenData.priceChange
  ];
  // Simple machine learning model to predict rug-pull risk
  const prediction = logisticRegression(features);
  return prediction > 0.5 ? "High Risk" : "Low Risk";  // Risk of rug-pull
}

Where:

features — token features (e.g., liquidity changes, unusual transactions), logisticRegression — logistic regression model for prediction.

2. Real-Time Alerts

⚠️ Instant notifications

Sends real-time alerts to users about risks associated with specific tokens or projects. Network activity monitoring for fast reaction to emerging threats. Alert Threshold:

function sendAlertIfSuspicious(transaction) {
  const deltaT = Math.abs(transaction.time - previousTransactionTime);
  const deltaP = Math.abs(transaction.price - previousTransactionPrice);
  const alertThreshold = 1.5;
  if (deltaT / avgTime > alertThreshold && deltaP / avgPrice > alertThreshold) {
    sendNotification("Suspicious transaction detected!");
  }
}

Where:

deltaT — change in transaction time, deltaP — change in transaction price, alertThreshold — threshold to trigger the alert.

3. Token Behavior Analysis

📊 Tracking token behaviors Analyzes historical token data to detect suspicious patterns. Identifies anomalies in tokens that may indicate fraudulent activities or malicious behavior.

function analyzeTokenBehavior(tokenHistory) {
  const avgPrice = getAveragePrice(tokenHistory);
  const anomalies = tokenHistory.map(price => Math.abs(price - avgPrice));

  const anomalyIndex = anomalies.reduce((sum, anomaly) => sum + anomaly, 0) / tokenHistory.length;
  return anomalyIndex > anomalyThreshold ? "Anomalous Behavior" : "Normal Behavior";
}

Where:

tokenHistory — historical prices of the token, avgPrice — average token price over the period, anomalyThreshold — threshold to flag anomalous behavior.