Roadmap & Vision: A Reusable Infrastructure Layer for Agent-Orchestrated Games
A protocol and benchmark for Claude Code–style agentic orchestration in games — not a demo, the runtime layer underneath.
// blog
Latest from the lab.
Thoughts on AI game design, world-building, and the future of interactive narrative.
WorldLines: Building a 🦞-style Agents Engine to Run TRPG
From AI chatbots to Claude Code TRPG, then building my own multi-agent engine. A four-month journey of a tabletop RPG game master.
Six Things Only WorldLines Does
Six things you'll feel in WorldLines that you won't find in chat-based AI roleplay. Less about architecture, more about what changes when AI characters actually live.
Version Control for Story Worlds — WorldLines' Commit-Tree
Today's AI roleplay evaporates the moment you close the tab. WorldLines makes worlds and souls version-controlled — a commit-tree you can branch, restore, and share.
Why We Start in Tokyo
What Ludic Dynamics builds, how it differs from existing AI roleplay, and why Tokyo is our launchpad.