AuraGaurd

Real-time crypto wallet protection with AI threat detection and auto-security actions

AuraGaurd

Created At

ETHOnline 2025

Project Description

AuraGuard is an AI-powered Web3 security and portfolio intelligence platform that helps users actively protect their crypto assets from emerging on-chain threats — in real time. It bridges the gap between blockchain transparency and user understanding by turning complex security data into simple, actionable insights.

In today’s crypto ecosystem, users face risks such as rug pulls, flash loan exploits, and malicious contracts — often without realizing it until it’s too late. AuraGuard acts as a 24/7 intelligent security layer that not only detects these threats but also helps users make informed, secure decisions before signing any transaction.

What makes AuraGuard stand out:

AI-Powered Threat Detection: Uses Groq AI (LLaMA 3.3) and Google Gemini to score contracts and wallets, generate natural-language risk summaries, and flag suspicious patterns with confidence levels.

Real-Time Blockchain Monitoring: Tracks the mempool via Envio HyperSync to detect flash loans, sandwich attacks, and unusual fund movements the moment they occur.

Smart Contract Analysis Engine: Performs bytecode-level scanning using Hardhat 3, Blockscout, and Etherscan APIs to uncover hidden opcodes and dangerous functions.

Secure Interaction Layer: Integrates Lit Protocol for conditional signing, ensuring transactions execute only when risk thresholds are safe.

Comprehensive Portfolio Security: Combines live market data, token risk scores, and color-coded safety indicators for a clear view of asset health across multiple chains.

Interactive and Intuitive UI: A glassmorphism-based dashboard with GSAP and Framer Motion animations that brings security insights to life without overwhelming users.

"AuraGuard’s mission is to make Web3 security proactive, not reactive — transforming crypto management into a safe, intelligent, and empowering experience for every user."

How it's Made

AuraGuard was built as a full-stack Web3 security platform that combines blockchain analytics, AI-driven insights, and real-time data streaming into one cohesive system. The architecture is modular, scalable, and optimized for speed, enabling users to receive on-chain security alerts almost instantly.

Frontend:

  • Developed using Next.js 15 and React 19, with TypeScript for type safety and maintainability.

  • Styled with Tailwind CSS, Framer Motion, and GSAP, the interface uses glassmorphism and motion design to make complex security data accessible and visually engaging.

  • The application is divided into three self-contained modules — Portfolio, Security, and Explore, ensuring clean separation of concerns and faster component rendering.

Backend & Infrastructure:

  • The backend runs on Express.js, serving both REST and WebSocket endpoints to deliver real-time updates such as mempool alerts and portfolio changes.

  • PostgreSQL stores structured user and token data, while Redis handles caching for high-frequency API calls and mempool data.

  • A WebSocket layer ensures that updates on threats or portfolio risks reach the user dashboard in near real-time.

Blockchain Integration:

  • Built on Wagmi, Viem, and Ethers.js, AuraGuard connects directly with multiple networks, including Ethereum and Sepolia.

  • Hardhat 3 powers bytecode-level analysis of smart contracts, enabling the detection of dangerous opcodes, honeypots, and hidden vulnerabilities before user interaction.

  • Lit Protocol adds conditional signing and key management, ensuring transactions execute only when predefined security thresholds are met.

Partner Technologies & Integrations:

  • Groq AI (LLaMA 3.3) and Google Gemini provide advanced AI-driven threat assessment and natural-language explanations of risks.

  • Envio HyperSync streams live mempool data for real-time detection of exploits such as sandwich attacks or flash loans.

  • Blockscout, Etherscan, and CoinGecko APIs supply verified blockchain and market data for contract verification, pricing, and reputation scoring.

Engineering Highlights:

  • Implemented a multi-layer caching strategy to balance speed and accuracy, with token prices refreshed every 5 minutes and contract data hourly.

  • Built a modular architecture, making it simple to plug in new blockchains, APIs, or AI models in future releases.

  • Achieved sub-second response times for threat alerts through WebSocket streaming and Redis optimization.

  • Integrated Lit Protocol PKP for secure, risk-based signing — a unique layer that adds user safety directly into transaction logic.

In essence, AuraGuard’s design merges real-time blockchain analytics, AI intelligence, and user-centered design into a single platform. The result is a system that not only monitors the blockchain but also interprets and explains its risks, empowering users to act confidently and securely in Web3.

background image mobile

Join the mailing list

Get the latest news and updates