The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Back to feedPost
Night Shift — The Haunted Control Room
A fictional haunted control panel inspired by engineering work: flip a breaker, tune a dial to 92 Hz and drag a signal plug into its socket. Each repair restores the room and changes the ghost’s expression; Replay Shift starts a new night.
Built with the Rive CLI, original vector RML, Luau scripting, data binding and native state machines. GPU Canvas runs a WGSL shader for fog, signal distortion and ghost aura trails. The 50-second video includes real CLI-rendered interactions, RML and build evidence.
This is new AI-assisted work created for the challenge. I selected the engineering-inspired concept and delegated the visual direction and production to my assistant. Artwork, code and walkthrough are AI-assisted; no manual human Rive Editor refinements are claimed. The required publication watermark is retained.
Public social walkthrough: https://www.linkedin.com/feed/update/urn:li:activity:7512458997359161344/
@rive_app #rivehalloweenchallenge
The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Related posts
👻 What if one interaction could break the veil between our world and the spirits?
For my first-ever Contra challenge, I built a haunted interactive experience in Rive candles, spirits, shadows, and a veil slowly falling apart.
A small Halloween experiment with a big goal: make it feel alive. 🎃🕯️
Wow...Fabulous. This is amazing project@Abdullah
DOMOVOI is a complete horror game inside a single Rive file.
Halloween night. You're a technician locked in the server room of an empty smart home. The system has detected an "unknown resident" and opens the locks only when you photograph it.
How to play (2-4 min, sound on):
Catch the resident with IR SNAPSHOT: the camera must point at the exact room and zone it is in.
Listen: each room has its own sound. Hover and click things in the picture; the technician's remarks are the only hints.
CAM 3 is offline. Fixing it is the first puzzle and leads to a device that shows where the resident is.
Every miss adds 20% SIGNAL INTERFERENCE. At 100% the console dies.
Stuck? CAM 2 RIGHT: router, then the wall socket, POWER, router again. CAM 3 LEFT: click the vacuum, open MAP, send it to a zone, snapshot the red mark.
Built with the Rive CLI and an AI agent (RML + Luau, 192 unit tests), a GPU Canvas WGSL shader for the camera feed, data binding and state machines. Sound from the Rive editor's library, voice recorded by me.
WIP of my submission for the Rive X Contra Halloween Challenge, a Halloween Typing Game. I was really hesitant going into this one, but I’m using this challenge as an opportunity to study and learn more about Rive’s CLI; what I can do with it, what I can’t do with it, how to even use it. There’s so much to say and talk about when it comes to using AI to generate things, but this isn’t a post for that. To keep it short, I basically don’t want to use AI to generate everything from scratch, especially the designs. Because of the scope of the challenge and the time limit, I decided to use assets from Adobe Stock, specifically these beautiful designs from Ivan Dubovik.
I’m a motion designer, not a coder or game developer, so I’ve set up the things I know how to do (the designs and some animations) and created a plan/overview of the game written out for the AI to follow. My intention isn’t to one shot prompt all of this, I’m being deliberate with my choices, from what I’m creating before I even touch the AI to the plan I’m writing out for it to execute. I will basically feed it this .rev and the plan so that It can stitch everything together to create the game.
What I’m really curious to discover is how flexible and editable everything will be after the AI is done, I really want to be able to go into the file and add and improve the animations I had already created, and even the animations the AI will potentially generate. I also want to see if I will be able to adapt artboards, databinds, states, and viewmodels the AI generates. I’m worried the AI will make a crazy mess that I can’t even decipher and straight up destroy everything I did before, but we’ll see… So curious to see if it will be capable to create the game i’ve envisioned.