Hey everyone,
I’ve been experimenting with designing an escape room experience built entirely around cards — something in the style of games like Unlock!, Deckscape, or Escape Tales — combined with a simple web app for additional interaction.
Recently, I’ve been trying to integrate AI into the design process (for puzzle generation, story building, and even creating card visuals). But I’ve run into a recurring challenge: AI tends to default to detective-style scenarios (investigations, evidence, crime scenes, etc.) whenever I mention “escape room.”
That’s not really what I’m aiming for.
What I want instead is a cohesive set of diverse puzzles — logic, codes, visual clues, etc. — all tightly connected to a central theme or narrative, without leaning into crime/investigation tropes. The goal is more about immersive storytelling and puzzle variety rather than solving a case.
I’ve noticed that getting useful results from AI requires very specific prompting and structuring — especially if I want:
•Puzzles that logically connect to each other and the story
•Thematic consistency across cards
•Visual assets that match the puzzle design
•Integration with a web app component
So I’m curious:
Has anyone here experimented with using AI to design escape room-style games (especially card-based ones)?
How do you guide the AI to stay within a specific theme and avoid unwanted tropes?
Any tips for structuring prompts or workflows to get better puzzle coherence?
Would really appreciate hearing about your experiences or any tools/methods you’ve found effective.
Thanks!