Mango
CT

XBotter Open Source

This was my first build of XBotter, to generate and post tweets with AI. It was hardcoded to spit out unfiltered tweets every hour about a random topic from an array that I predetermined.

Loading stats...
Visit Website

Tech Stack

  • • Node.js & Javascript
  • • X API Integration
  • • AI Language Models

AI Integration

  • • Created an OpenAI Api Key
  • • Integrated the Api Key for tweet generation

⚠️ Development Challenges

I was still trying to get my feet wet in the twitter community, therefore I didn't really know if my bot's tweets were good or not. I refined the prompt multiple times, even tried different text completion + generation LLMs from Anthropic and Mistral. I ended up preferring OpenAI, and just kept posting the generated tweets that I liked, on a schedule.

Documentation & Steps

XBotter Documentation

Development Process

I wanted this build to be very user friendly when setting up, and can be easily manipulated with newer prompts, better LLMs. That way I can just swap out the API keys whenever the company's stock goes down.. just kidding 😅

Core Features

  • Modular structure for handling different LLM API keys
  • Utilizes Cron-Job for scheduling tweets according to user desire
  • Can be preset with an array of keywords and sentences or topics, and then configured from an async function
  • Thanks and References

    Inspired by Ryan Carmody's tutorial and Gabe's support

    Disclaimer

    This is an experimental project showcasing AI capabilities. Use responsibly and in accordance with X's Terms of Service.

    ← Go back

    Let's ConnectLet's ConnectLet's ConnectLet's ConnectLet's ConnectLet's ConnectLet's ConnectLet's ConnectLet's ConnectLet's ConnectLet's ConnectLet's Connect