DEV Community

Cover image for The Last Programmer 👻 — A Haunted AI Coding Assistant Powered by Local LLMs
Shubham Bawari
Shubham Bawari

Posted on

The Last Programmer 👻 — A Haunted AI Coding Assistant Powered by Local LLMs

Hacktoberfest Open-Source AI Challenge: Ghost in the Machine Submission 👻🌼

👻 The Last Programmer — A Haunted AI Coding Assistant

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 4: Ghost in the Machine

What I Built

👻 The Last Programmer is a spooky AI coding assistant powered by open-weight AI.

It brings an ancient programming spirit to life. Users can chat with the ghost to learn programming concepts, understand code, and get help debugging errors.

Features

  • 👻 Three ghost personalities: Friendly Ghost, Evil Ghost, and Ancient Programmer.
  • 💻 Programming explanations and debugging assistance.
  • 🟣 Haunted purple-and-black interface.
  • 🔒 Local AI inference using Ollama.
  • 🌐 Offline support after initial setup and model download.

Code

🔗 GitHub Repository: The Last Programmer

How I Built It

I built this project using:

  • Python — application logic.
  • Streamlit — interactive web interface.
  • Ollama — local AI inference.
  • Qwen2.5-Coder 3B — open-weight coding model.
  • Requests — communication with the local Ollama API.

The application sends prompts to a locally running AI model. A system prompt defines the ghost's personality while keeping its programming explanations useful.

The model runs locally instead of relying on a hosted AI API.

Run Locally

Step 1 — Install Ollama

Download it from ollama.com.

Step 2 — Download the model

ollama pull qwen2.5-coder:3b
Enter fullscreen mode Exit fullscreen mode

Step 3 — Install dependencies

pip install -r requirements.txt
Enter fullscreen mode Exit fullscreen mode

Step 4 — Launch the application

streamlit run app.py
Enter fullscreen mode Exit fullscreen mode

Why Does Open Innovation Matter?

Open-weight AI gives this project flexibility and control.

  • Privacy: Prompts are processed by the local Ollama server rather than sent to a hosted AI API.
  • Accessibility: No paid inference API is required.
  • Customization: The model and system prompt can be changed.
  • Offline capability: The application can work without internet after setup and model download.

Open innovation lets developers experiment, customize AI behaviour, and build tools without depending entirely on closed services.

The Spirit of It 👻

What if an old programmer never really disappeared?

What if forgotten programming languages and abandoned code could speak again?

The Last Programmer imagines a ghost living inside a computer, ready to help the next generation of developers.

Code may grow old. Knowledge should never die.

Top comments (0)