StarkFinder: Your AI-Powered Starknet Navigator

Table of Contents

• Introduction
• Features
• Architecture
• Getting Started
  • Prerequisites
  • Installation
• Using StarkFinder
  • Available Agents
    • Trade Agent
    • Investment Agent
    • Exploration Agent
    • Research Agent
• Design
  • Landing Page
  • Starkfinder Transaction Page
• Future Plans
• Contributing
• License

Introduction

StarkFinder is an innovative AI-powered application that helps users navigate the Starknet ecosystem with ease. Developed for the EthGlobal AI Web3 Hackathon, StarkFinder leverages the power of multi-agent AI to automate various tasks, making the Starknet experience more efficient and streamlined for users.

Features

• Multi-Agent Architecture: StarkFinder supports a diverse set of specialized agents, each with its own capabilities, such as trading, investing, exploring, and researching Starknet protocols.
• Starknet Integration: The agents seamlessly interact with the Starknet Layer 2 network, allowing users to leverage the benefits of this high-performance blockchain.
• Conversational Interface: Users can communicate with the agents through an intuitive chat-like interface, making the experience natural and user-friendly.
• Automation and Task Delegation: Users can delegate tasks to the agents, which will then execute them on their behalf, saving time and effort.
• Comprehensive Starknet Insights: The application provides users with in-depth information about Starknet, including DeFi platforms, trading opportunities, and research insights.

Architecture

StarkFinder is built using the following technologies:

• Next.js: A React framework for building server-rendered and static websites.
• Prisma: A type-safe ORM for interacting with the database.
• PostgreSQL: The relational database used for storing user data and agent configurations.
• TypeScript: A superset of JavaScript that adds optional static typing.
• Cairo: A domain-specific language used for writing Starknet smart contracts.

The application's architecture follows a multi-agent design, where each agent specializes in a specific task or domain. The agents communicate with the Starknet network through a set of APIs and integrate with various Starknet-based protocols and services.

Getting Started

Prerequisites

• Node.js (version 14 or higher)
• PostgreSQL (version 12 or higher)
• Create .env file both in client and tg_bot directories with different DATABASE_URL. For BrianAI API key go to their app connect wallet and get your key.

Installation

1. Clone the repository:

   git clone https://github.com/username/StarkFinder.git
   

2. Navigate to the project directory:

   cd StarkFinder
   

3. Install the dependencies:

   cd client
   npm install
   cd ..
   cd tg_bot
   npm install
   

4. Set up the database (both into client and tg_bot directories):

   npx prisma generate
   npx prisma migrate dev
   

5. Start the development server:

   npm run dev
   

Setting up Telegram Test Environment

• Please go through this doc to setup the telegram mini app test environment

Using StarkFinder

Available Agents

Trade Agent

The Trade Agent is responsible for executing trades on the Starknet network based on user-defined conditions. Users can interact with this agent using the /trade command.

Investment Agent

The Investment Agent researches and automatically invests in Starknet-based decentralized finance (DeFi) platforms on the user's behalf. Users can interact with this agent using the /invest command.

Exploration Agent

The Exploration Agent helps users discover new Starknet-based projects, protocols, and services. Users can interact with this agent using the /explore command.

Research Agent

The Research Agent provides in-depth analysis and insights about the Starknet ecosystem. Users can interact with this agent using the /research command.

Design

Landing Page

Kindly refer to the following for the landing page design.

Starkfinder Transaction Page

Kindly refer to web app for the web application transaction page.

Future Plans

In the future, the StarkFinder team plans to expand the application's capabilities by integrating with more Starknet-based protocols and services. Additionally, the team aims to explore the integration of XMTP (Cross-Message Transport Protocol) to enhance the decentralized and secure communication between the agents and users.

Contributing

Contributions to the StarkFinder project are welcome. If you'd like to contribute, please follow these steps:

Refer to Contributing for contributing.

License

This project is licensed under the MIT License.

Contributors