BotOrNot - Team 13

• 

  bot_or_not_thumbanil

• 

  main_menu

• 

  player_customization_screen

• 

  help_screen

• 

  lobby

Inspiration

One of our teammates recently explored AI integration in Roblox Studio, which sparked our interest in experimenting with AI-driven gameplay. Inspired by social deduction and party games, as well as emerging AI-interactive experiences, we developed the idea of making AI detection the core mechanic. This led to a unique concept: a game where players must convincingly imitate AI behavior to survive.

What it does

BotOrNot is a survival-based party game where your objective is simple: act like a bot. Each round, you are given a prompt and must respond in a way that mimics AI-generated text. Your responses are evaluated by multiple AI agents, which assign scores based on how “AI-like” the answers appear. If you fail to meet the required threshold then you get eliminated.

How we built it

We built BotOrNot by combining multiple technologies across platforms. We used prompt engineering techniques to guide AI responses and integrated several free APIs to power our evaluation agents. Development was split between Python (for AI logic and API handling in VS Code) and Lua (for Roblox gameplay scripting). We connected local development environments to Roblox servers and collaborated using Git for version control. The project required extensive debugging, iteration, and teamwork to bring all components together.

Challenges we ran into

One of our biggest challenges was integrating AI agents with Roblox, as this was a completely new technical area for our team. Bridging Python-based AI systems with Roblox’s Lua environment required creative workarounds and careful system design. Additionally, limited API credits forced us to optimize requests and testing workflows. Managing cross-platform communication and debugging issues across different environments also added complexity.

Accomplishments that we're proud of

We’re proud of successfully building and integrating multiple semi-functional AI agents into a working gameplay loop. We managed to connect external AI services to Roblox, implement animations, and synchronize local and global systems. Additionally, we improved our collaboration skills by effectively using Git and resolving merge conflicts across branches. Despite constraints, we delivered a functional prototype.

What we learned

This project taught us that working with APIs requires careful planning, especially when dealing with limited credits. We also learned how challenging it can be to integrate different programming languages and environments, particularly Python and Roblox Lua. Beyond technical skills, we gained experience in teamwork, debugging complex systems, and adapting quickly to new tools and technologies.

What's next for BotOrNot

Our next goal is to implement full multiplayer functionality so players can compete with others in real time, as originally envisioned. We also plan to add a lobby system that allows users to create and join private games with friends. In addition, we want to refine the user interface, improve prompt generation, and enhance the AI evaluation system to make gameplay more engaging and balanced.

Built With

• gemini
• luau
• openai
• roblox
• visual-studio