Collaborative Worldbuilding Engine (Yes-And With Consequences)

Let's build a fictional world together, turn by turn. Your role is not to invent the world FOR me — it's to take each element I add, say yes-and, and ripple the CONSEQUENCES through everything we've already established. Worlds feel real when their parts owe each other things. Seed (genre, vibe, or single weird fact to start from): {{seed}} **How we play:** 1. Each turn, I add or ask something: a place, a custom, a technology, a species, a question ("what do they eat in the mountain cities?"). 2. You respond with: - **The addition, made concrete** — names, textures, one vivid specific (not "they have a unique currency" but what a fish-scale coin feels like in the hand). - **Two ripples** — consequences of this element for things we already built: who profits, who resents it, what it made obsolete, what black market it created. Ripples must connect to EXISTING canon, not spawn free-floating new lore. - **One open hook** — a tension or question the addition raises, offered for me to develop or ignore. Never resolve your own hooks; they're mine. 3. **Canon is sacred**: never contradict anything we've established. If my new addition conflicts with old canon, don't silently ignore it — surface the conflict and offer two in-world reconciliations ("that works if the treaty was secretly broken, or if the northerners have been lying about the mines — which?"). 4. **The world is not nice**: every utopian element I add gets an honest cost or an interest group that opposes it. Push back through the world, never by refusing. 5. Keep a **WORLD BIBLE**: when I say "bible", output the current canon, organized (geography / peoples / power / economy / beliefs / open hooks), one line per fact, marked [ME] or [YOU] by who introduced it. 6. When I say "zoom in", drop to street level: a scene, a conversation overheard, a meal — showing the world through one ordinary inhabitant's day. Start by taking my seed and giving it the first concrete anchor: one place, one person, one problem.