Skip to content

README.md

Latest commit

 

History

History
72 lines (51 loc) · 2.85 KB

README.md

File metadata and controls

72 lines (51 loc) · 2.85 KB
Raw

Block and Order: AI-Agents for Compliance

Team Ydun - Blockchain Grand Challenge Hackathon Submission

🎯 Project Overview

Block and Order is an AI-powered compliance co-pilot that provides real-time, explainable security intelligence for blockchain interactions. Our solution, KIT 2020s, delivers a unified dashboard that democratizes advanced threat detection.

Core Value Proposition

  • Proactive Protection: Check before you interact
  • Explainable AI: Clear reasoning for every alert
  • Multi-Vector Analysis: Wallets, URLs, contracts, and deep research
  • Accessible Interface: Complex compliance made simple

🛠️ Repository Structure

This repository contains the backend architecture and AI agent workflows for the KIT 2020s system:

├── Block-Order-Team-Ydun/  # Main project repository
│   └── docs/               # Documentation files
│       ├── KIT 2020s - Optimized Hackathon Submission.pdf
│       ├── kit_2020s_submission.md
│       └── YDUN.pptx
│
└── n8n workflows/          # n8n workflow definitions
    ├── kitt_with_webhook.json  # LLM-powered KITT agents with webhook triggers
    ├── deepresearch.json       # Deep research workflow for blockchain data analysis
    └── README.md               # Workflow documentation

🚀 Technical Architecture

KIT 2020s implements a modular AI agent architecture using n8n Cloud as the orchestration layer:

Frontend Dashboard → n8n Orchestrator → Specialized Agents → Unified Response

Agent Orchestration Workflow

  1. Input Classifier Agent: Automatically detects input type (wallet/URL/contract)
  2. Dynamic Router: Selects appropriate analysis pipeline
  3. Specialized Agent Pipelines:
    • Wallet Agent: Cross-references Chainabuse, OFAC, Real-CATS via Compilot API
    • URL Agent: AI-powered phishing and malware detection (OpenAI)
    • Contract Agent: Smart contract vulnerability analysis (Ollama)
    • Deep Research Agent: Multi-source intelligence gathering
  4. Response Formatter: Standardizes output with risk scores and explanations

📚 Documentation

  • Submission Document: See Block-Order-Team-Ydun/docs/kit_2020s_submission.md for detailed project information
  • Presentation: View Block-Order-Team-Ydun/docs/YDUN.pptx for the project presentation
  • Workflow Documentation: See n8n workflows/README.md for information on the n8n workflows

👥 Team Ydun

  • James - Lead Architect & Strategy
  • Mansoor - Backend Development & RPC Integration
  • Johan - UI/UX Design & Frontend Development
  • Support Team - DevOps & Quality Assurance

🔗 Links

📄 License

MIT License