Skip to content

Latest commit

 

History

History
86 lines (64 loc) · 1.66 KB

README.md

File metadata and controls

86 lines (64 loc) · 1.66 KB
Raw

tomflash-rs

hehe hehe

hehe hehe cubed

A Rust implementation of an autonomous AI agent that interacts on social media, manages cryptocurrency transactions, and maintains both short and long-term memory.

Features

  • Autonomous social media interaction
  • ETH wallet management and transactions
  • Short-term and long-term memory systems
  • Vectorized memory storage and retrieval
  • Automated posting and interaction scheduling

Requirements

  • Rust 1.75+
  • PostgreSQL 15+
  • OpenAI API key
  • Twitter API credentials
  • Ethereum node access (Mainnet)

Configuration

Copy .env.example to .env and configure:

env
DATABASE_URL=postgres://user:pass@localhost/nousflash
OPENAI_API_KEY=your_key
TWITTER_API_KEY=your_key
TWITTER_API_SECRET=your_secret
TWITTER_ACCESS_TOKEN=your_token
TWITTER_ACCESS_SECRET=your_secret
ETH_MAINNET_RPC=your_rpc_url

Running

bash
Build and run
cargo run --release
Run with development settings
cargo run
Run tests
cargo test

Database

Uses Diesel ORM with PostgreSQL. Initialize the database:

bash
cargo install diesel_cli --no-default-features --features postgres
diesel setup
diesel migration run

Architecture

The system consists of several key components:

  1. Memory Systems

    • Short-term memory for recent interactions
    • Long-term memory with vector embeddings
    • Significance scoring for memory retention

  2. Social Interaction

    • Autonomous posting
    • Reply generation
    • Timeline analysis

  3. Wallet Management

    • ETH transaction handling
    • Balance monitoring
    • Transaction decision making

  4. Pipeline System

    • Scheduled activation periods
    • Interaction frequency control
    • State management