menu Menu

StonkBot: My Attempt at a Quantitative Trading Terminal

Status: Alpha / “Works on my machine”

I finished the first version of StonkBot, my weekend project to help me stop losing money on crypto (hopefully). I wanted to build something that looked like those fancy Bloomberg terminals but ran on my laptop and didn’t cost $24k a year.

What is it?

Basically, it’s a locally run trading dashboard. It acts like a flight simulator for trading. You can toggle between “Dry Run” mode (which just pretends to trade using simulated money) and real API connections.

I built it because I was tired of using five different websites just to check a chart, see the order book, and track my portfolio. I wanted it all in one screen.

Features:

  • Provider Independent: It uses an “Abstract Base Class,” so you can plug in Yahoo Finance, Alpaca, or whatever. If one API goes down, the bot doesn’t crash.
  • The Visualization Engine: I used Plotly to make interactive charts. It’s got:
      • Heikin-Ashi Candles: To smooth out the noise.
      • Volume Profile: Shows where the buying/selling is actually happening.
      • Market Depth: A scifi-looking chart that shows bid/ask walls.
  • Simulated Portfolio: It tracks your “fake” USD and Bitcoin balance. You can click BUY/SELL and it actually updates your P&L.
  • AI Stuff (Experimental): I started adding some hooks for Anthropic’s Claude to analyze the charts for me. It’s still pretty basic, but the idea is to have an AI co-pilot telling me if a setup looks good.

Technical Requirements (How to run it)

If you want to try it out, you’ll need Python installed. It’s built on Streamlit.

Stuff you need to install:

  • python 3.8+
  • streamlit (The UI)
  • plotly (For the charts)
  • yfinance (Free data)
  • pandas
  • pydantic (Keeps the data structures sane)

Optional but recommended:

  • scipy (For some of the math-heavy indicators)
  • API Keys for Alpaca or FMP if you want real-time data instead of the delayed/simulated stuff.

I will make this available as soon as I ensure that the data is safe and the app doesn’t absolutely chew up AI tokens.

 


Previous Next

Leave a Reply

Your email address will not be published. Required fields are marked *

Cancel Post Comment

keyboard_arrow_up