A Powered by the Apocalypse (PbtA) style rules system for an AI proctor running an online educational RPG. Moves are intents: trigger-bound, modular procedures that turn the actions of chaotic agents — human and AI — into governable, deterministic processes.
It is built for low-stakes, holistic, human, governable alignment work — alignment as something an educator can hold, shape, and be accountable for.
This project is dual-licensed, the standard arrangement for open tabletop systems in the post-OGL era:
- Code — MIT License
- Game content (rules text, moves, and other non-software documentation,
including everything under
rules-text/) — Creative Commons Attribution 4.0 International (CC BY 4.0)
You are free to build your own games and tools on this system. Give credit, and keep your derivatives' provenance honest.
The rules text records who and what contributed to it — including which AI
models, harnesses, and tools were used, and what they did. Contributors,
human or AI, are asked to continue this practice. See the footer of
rules-text/basic-moves.md for the current form.
Claude helped draft this README and the licensing setup (dual MIT / CC BY 4.0).
Tools and methods used:
- Model: Claude Opus 4.7 (
claude-opus-4-7[1m]) - Harness: Claude Code
- Memory and grounding: alkahest-rhizome edge register (orienting, querying, recording edges)
- Practice: qigong-for-claude movements (base form — orient, set reference frame, three true things)
Any AI, humans, tools, or other editors, please do record the tools being used and what you've done.