wheelgpt

More information on wheelgpt.dev. Get the OpenPlanet Plugin.

WheelGPT is a Trackmania Twitch bot that lets viewers interact with the stream in real time. Viewers can use chat commands to check the current map, guess the streamer’s next personal best time, and get information about the current Trackmania room the streamer is playing in. All in-game data is collected and sent to the server by a dedicated Openplanet plugin, ensuring the bot always provides up-to-date information. Originally developed for GranaDyy – the wheel guy – WheelGPT now brings interactive Trackmania experiences to any Twitch channel.

The frontend application, built with Angular and available at wheelgpt.dev, provides an intuitive interface to configure the bot and connect it to your Twitch channel. The web application features support for both light and dark themes, and offers multilingual support for German, French, and English.

See the Screenshots section below for visual examples of the interface.

Any Questions? Feedback?

For more detailed information, check out the documentation.
If you have any questions regarding the plugin or the Twitch Bot, feel free to contact me:

• Discord: sowiemarkus
• [email protected]

Features

Token

To use this plugin, you need a token. For that you need to register on wheelgpt.dev. The token assigned to you is sent with every request to our server and identifies your Twitch channel. So don’t share this token with anyone. Otherwise, others can use the plugin under your name and, for example, send best times to your Twitch channel that you have never driven.

The token assigned to you must be entered in the Settings -> WheelGPT-Plugin -> Info/Token.

Personal Best

With the Twitch Bot, your viewers can guess what your next best time will be with !guess <time>. The format of the time is hours:minutes:seconds.milliseconds. Hours and minutes are optional. Example: !guess 12.345 or !guess 01:23.456.

If the plugin is activated, whenever you set a new best time on a map, this time will be sent to our server. Afterwards all received guesses of your users will be checked and in the bot will announce in the chat which viewer was closest with the guess.

If you want to deactivate that your best times are sent to our server, you can deactivate this under Settings -> WheelGPT-Plugin -> Send PBs to Server.

WheelGPT features a leaderboard that tracks viewer performance across guessing sessions. The leaderboard is accessible at https://wheelgpt.dev/leaderboard/{channelId} and provides:

Scoring System:

• Perfect Guess: 5 points for guessing the exact time
• Closest Guess: 1 point for being the closest in a round

Features:

• Public Rankings: View the top guessers in each channel
• User Search: Find specific users and their current position
• Paginated Results: Browse through all participants
• Real-time Updates: Leaderboard updates automatically as new scores are recorded

Viewers can use the !lb command in chat to get a quick link to the leaderboard, or use !lb <username> to check a specific user’s position and points.

The leaderboard encourages friendly competition among viewers and adds a persistent progression element to the guessing game, making each stream more engaging for regular participants. (And more YEK ofc)

Map

With the Twitch bot, viewers can use the command !map to find out which map you are currently playing.
If the plugin is activated, whenever you enter a new map, this map is sent to our server. If you use the Champion Medals plugin, the Champion Medal time will also be sent to the server.

If you want to disable this feature, you can deactivate this under Settings->WheelGPT-Plugin-> Send Maps to Server.

Room

With the Twitch bot, viewers can use the command !room to get information about the current Trackmania room the streamer is playing in.
If the plugin is activated, whenever you join a room, the room information (such as name, host, and player count) is sent to our server. This allows viewers to always see up-to-date room details via the bot.

If you want to disable this feature, you can deactivate it under Settings -> WheelGPT-Plugin -> Send Rooms to Server.

Getting Started

1. Create a .env file based on the provided example.

2. Build the Docker image:

   docker build -t wheelgpt-app .
   

3. Run the Docker container:

   docker run -p 3000:<port_from_env> wheelgpt-app
   

   This will start the backend server on port 3000.
   You can adjust the port mapping as needed.

Development

1. Run npm run setup
2. Run npm run dev:all

Screenshots