Solana AI Agent

Overview

Why Choose Solana AI Agent?

Comprehensive Features

• Built on CyberChipped: A robust Python framework for OpenAI assistants.
• Enhanced Capabilities: Offers conversational memory, parallel function execution, smart tool selection, and message history with MongoDB.
• Frameworks Utilized: Leverages FastAPI and Next.js, two of the most popular web frameworks.
• Customization Made Easy: Quickly integrate custom functions into your AI agent with minimal code.
• Key Actions:
  • Transfer tokens from the Agent wallet to other wallets via Helius.
  • Swap tokens within the Agent wallet using Jupiter and Helius.
• Social Media Integration:
  • Connects to X through the X API (Basic Plan).

Superior to Eliza

• No Code Changes Required: Just add simple environment variables along with MongoDB/Redis for an out-of-the-box AI agent with real-time chat.
• Advanced Conversational History: Outperforms Retrieval-Augmented Generation (RAG) in user interaction, tool usage, and memory/context recall.
• Efficient Tool Usage: Offers parallel tool calling and automatic AI tool selection, surpassing any LLM completion API.
• Simplicity in Design: An opinionated framework with a straightforward approach to functionality.
• Python-Based: Developed in Python, the leading language in AI and GitHub projects.
• Real-Time Responses: Interacts on X in real-time, responding only when necessary.

Local Development

Running Locally on Mac or Linux

1. Clone the repository:

   git clone https://github.com/rizzolib/Solana-AI-agent

2. Ensure you have the latest LTS version of Node installed with Yarn.
3. Install Python 3.12.7 using Poetry.
4. Install Docker and Docker Compose.
5. Start the services:

   docker-compose up -d

6. Rename .env.sample to .env in both site and agent directories.
7. Set the PRIVATE_KEY (in base58 format) in the .env file within the site folder for Solana actions.
8. Obtain and set your OPENAI_API_KEY in the agent folder's .env file. Get OpenAI API Keys.
9. Set HELIUS_API_KEY and HELIUS_RPC_URL in both .env files. Learn More About Helius.
10. For XBot integration, create an X developer account and configure the keys in the agent folder's .env file; uncomment the relevant code in main.py.
11. Ensure all secrets match between the .env files, using uuidv4 or other secure keys.
12. Open two terminal windows:
    • Terminal 1:

      cd site && yarn install && yarn dev

    • Terminal 2:

      cd agent && poetry install && bash ./dev.sh

13. Access the application in your browser at http://localhost:3000.

Deployment

Deploying to Heroku or Dokku

1. Provision MongoDB and Redis databases.
2. Acquire a domain with two subdomains for the site and agent.
3. Configure the necessary environment variables on Heroku or Dokku.
4. Initialize Git repositories for both site and agent folders.
5. Commit and push changes to the main branch for both folders.
6. Scale the worker and scheduler to 1 (the web should already be set to 1).

Advanced Topics

Agent Functions

• Use clear snake-case naming for functions and descriptive parameter names.
• Functions should accept only str parameters and return a str.
• Ensure functions are synchronous; avoid async libraries or methods (e.g., use requests instead of httpx).
• Limit the size of tool outputs (strings) to accommodate multiple parallel calls within size limits.
• Consider the 128k token input limit when processing data, especially from APIs.
• Opt for OpenAI gpt-4o for better performance over gpt-4o-mini, unless cost is a concern.

Production Applications

• WB AI Agent
• Solana Agent X Bot