GhostScript AI：由本機 AI 驅動的鬧鬼終端

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

What I Built 👻

I built GhostScript AI, a spooky but helpful coding companion powered by local AI.

Users can ask questions about terminal commands and coding concepts. The ghost responds with spooky commentary, a beginner-friendly explanation, and a useful example.

I chose the haunted-machine theme to make learning programming more fun and less intimidating for beginners.

Code

GitHub repository: https://github.com/DhanushRajUrs/ghostscript-ai

How I Built It

My technology stack:

• Python
• Streamlit
• Ollama
• Gemma 3:1B

The user enters a question, chooses a spooky personality, and clicks Summon the Ghost. The application sends the prompt to Gemma through Ollama, which runs locally on my machine.

The model generates a playful response alongside a practical explanation.

Why Does Open Innovation Matter?

Using an open-weight model with local inference gives developers more control over their applications and the prompts they process.

Once the software and model are installed, GhostScript AI can generate responses without depending on a hosted AI API. This reduces reliance on external services and allows developers to experiment with local models.

The project also demonstrates how open AI tools can be used to build creative, interactive applications with a small technology stack.

The Spirit of It 👻

I wanted to combine the fun of a haunted computer with the practical goal of helping beginners learn programming.

Instead of making terminal commands feel intimidating, GhostScript AI turns coding questions into a playful conversation with a digital ghost.

This is an early prototype, and future improvements could include a terminal-style interface, more ghost personalities, and interactive coding lessons.

Building this project helped me practise Python, Streamlit, GitHub, and local AI integration.

Happy Hacktoberfest! 🎃