Pharos Devent: Decentralized AI Agent Marketplace

Pharos Devent is a next-gen AI x Web3 platform enabling users to build, own, and monetize AI agents as NFTs on the EDU-CHAIN blockchain. This document outlines the architecture, tech stack, and core features of the platform.

🔗 Smart Contract Addresses

• DeventAgent Contract:

  0xe886CC039Abe0184B807Dad82459B0C365a03fD8

• DeventToken Contract:

  0x8de73B26c229FcD50DEb8ce26c86Ebb6BC75Dbd9

⚙️ Core Features

• No-Code AI Agent Builder
  Define agents using natural language—Pharos auto-generates logic for deployment.

• Advanced IDE for AI Logic
  Monaco-based IDE for Python development with live preview and AI code assist.

• NFT-Based Ownership
  Each agent is minted as an ERC-721 NFT, with logic and metadata stored on IPFS.

• Tokenized AI Ecosystem
  Agents can issue ERC-20 tokens to incentivize use and interaction.

• Decentralized Execution
  Agents execute autonomously via smart contract triggers.

🧱 Tech Stack

Frontend

• Next.js (SSR + ISR)

• TypeScript

• Tailwind CSS

• Dynamic Routing for agent-specific pages

• Optimized NFT Image Previews via Next.js

• Server Actions/API Routes for AI logic handling

Web3 Integration

• EDU-CHAIN Testnet (Ethereum Layer 3)

• Privy: Seamless wallet connection and session management

• Wagmi + Viem: React hooks for wallet and contract ops

• OpenZeppelin ERC-721/20: Standards for NFTs and agent tokens

• Pinata + IPFS: Decentralized metadata and logic storage

🤖 AI Development & Security

• On-Chain AI Triggering via NFT metadata

• Monaco Editor with:

  • Python/Solidity syntax highlighting

  • AI-generated code snippets

  • Live agent behavior previews

• Lit Protocol: Encrypts agent logic and metadata

• Pinata (IPFS): Ensures integrity and permanence

• NFT-Based Access Control: Only owners can decrypt and run agents

🚀 Workflow

1. Create Agent
   Use no-code input or Monaco IDE.

2. Mint NFT
   Metadata is encrypted (Lit) and pinned to IPFS.

3. Deploy & Run
   AI logic is executed via smart contract on EDU-CHAIN.

4. Monetize with Tokens
   Each agent can launch its own ERC-20-based mini economy.

🧺 Experimental Features

• Encrypted AI Logic Execution via Lit Protocol

• ZKML (Zero-Knowledge ML): Research ongoing for trustless on-chain inference

• Programmable AI Tokenomics:

  • Agent NFTs mint a capped ERC-20 token supply

  • Subscription-based token distribution mechanisms

🗸 Roadmap

• ✅ AI Agent NFT Minting

• ✅ Tokenized AI Economy

• ✅ Web3 Wallet Onboarding

• 🔜 Full On-Chain ZKML Execution

• 🔜 Cross-Chain AI Marketplace Integration

🔐 Security & Reliability

• Lit Protocol ensures AI logic confidentiality

• Arbitrum Rollups provide transaction security and scalability

• Autonome Protocol enables decentralized AI execution with no single point of failure