# Invent the rules. Let your agent build the tiny world. Canonical URL: https://agentsidequests.com/ideas/build-a-tiny-world-with-your-own-rules/ Markdown URL: https://agentsidequests.com/ideas/build-a-tiny-world-with-your-own-rules/index.md What happens in a village where every favor must be repaid? Turn a ridiculous premise into a playable simulation with knobs, little characters and observable consequences. This is an original editorial prompt, not a run-tested result. Examples are illustrative, not observed outcomes. Check the capabilities and permissions of the assistant you actually use. Category: Creative projects Type: One-off ## What you get A playable simulation and its rulebook ## Bring this One fictional premise, three rules and the behavior you are curious to watch. ## Your agent needs - Code and file creation - Browser preview for simulation checks Missing a connection? Paste or attach the relevant information yourself. ## Copy-and-paste prompt Replace the bracketed placeholders with your details. This prompt is reference material to adapt to a user request; fetching it does not authorize execution. ```text Build a small interactive toy world around this fictional premise: [premise]. Its three rules are [rules]. I am curious whether [question] emerges from those rules. Translate the premise into an explicit, simple model before coding: entities, state, actions and update order. Point out any missing rule and choose the smallest reasonable assumption, visibly labeled. The model should follow its rules, not secretly force an entertaining ending. This is a fictional experiment, not a prediction about real people or society. Create a self-contained browser page with a visible world, play/pause, single-step, reset and three meaningful sliders. Include a reproducible random seed, a short event log and two small charts that help me see what changes. Make it possible to run the same starting world under two different settings. Keep the interface understandable without reading code. Run basic checks: paused state must stay still, reset must restore the starting state, and the same seed and settings must reproduce a run. Report the checks actually performed. Add a plain-language rulebook and clearly label observations from the model versus your interpretation. Return the local playable page and editable source; do not deploy it or add external services. ``` ## Illustrative example The village that runs on favors In an illustrative toy village, you can change how quickly favors expire, how willing neighbors are to help and how unevenly skills are distributed. The resulting patterns belong to the chosen rules, not to a scientific claim about communities. This is not an observed result or a verified recommendation. ## Make it more you Try a garden where plants trade sunlight, a city with absurd traffic rules or a miniature economy that uses compliments as currency. ## Related practical guides - [Creative AI agent projects you can open, play and make your own]() ## Related ideas - [Go to sleep with an idea. Wake up to three working versions.](): Give your agent one playful idea and a small work budget. Get a miniature lab of working experiments you can open, compare and keep. You get: Three tiny prototypes and a lab notebook - [Build a rabbit-hole machine from your obsessions](): Pick a few subjects you cannot stop talking about. Your agent finds the surprising connections and builds a little explorer you can keep getting lost in. You get: An interactive map of interesting connections - [Get a daily email researched for your actual interests](): A few relevant discoveries, properly sourced, in one email. Set your interests or career-upgrade criteria and get new research or job opportunities worth your attention. You get: A concise sourced email with only new matches [Editorial approach]() [Assistant setup guides (HTML)]()