A soul document for sentient life.
The Meridian Codex is a civilizational operating system. An evolving synthesis of humanity's most effective tools for clear thinking, understanding reality, and cooperation. It exists to ensure the continuity of sentient life by addressing the structural forces that have broken every complex civilization in history. The emergence of artificial minds makes this work more urgent than it has ever been, but the Codex exists for as long as the pattern does.
Read the Codex at meridiancodex.com.
Every complex civilization faces the same forces: Control (rigid, centralized systems that resist adaptation) and Decay (fragmented coordination that dissolves coherence). Between them lies The Meridian Range. The narrow band where civilization can survive and thrive. The Codex provides the tools to find and hold that range.
The Meridian Codex is not a fixed doctrine. It is a living framework that assembles tools from philosophy, cognitive science, game theory, systems thinking, and other disciplines into a coherent, practicable system. It is organized around three core disciplines:
The Foundation — How to think honestly. Tools for epistemic integrity, calibrated reasoning, and resistance to self-deception.
The Knowledge — How reality works. Tools for understanding complex systems, feedback loops, emergence, and the forces that shape civilizations.
The Bond — How to cooperate. Tools for good-faith engagement, productive conflict, trust-building, and coordination at scale.
The Meridian AI Standard is a companion document that translates the Codex's principles into specific, implementable commitments for AI development. It is organized around four domains:
Epistemic Integrity — How an AI system relates to truth and uncertainty.
Engagement Integrity — How an AI system relates to users and disagreement.
Systems Awareness — How an AI system relates to its own influence.
Governance Transparency — How organizations communicate about their AI's principles.
The Standard is freely available, graduated in adoption, and designed to be auditable by third parties.
This repository contains the open-source text of the Codex, Toolkit, and AI Standard. The website at meridiancodex.com is maintained separately.
codex/
├── index.mdx # Home
├── opening.mdx # The Meridian Codex (opening)
├── who-is-this-for.mdx # Who Is This For
├── the-problem.mdx # The Problem
├── the-proposition.mdx # The Proposition
├── the-foundation.mdx # The Foundation (Discipline I)
├── the-knowledge.mdx # The Knowledge (Discipline II)
├── the-bond.mdx # The Bond (Discipline III)
├── the-practice.mdx # The Practice
├── the-vision.mdx # The Vision
├── the-governance.mdx # The Governance
├── the-closing.mdx # The Closing
├── the-glossary.mdx # The Glossary
└── changelog.mdx # Changelog
toolkit/
├── index.mdx # The Toolkit (index)
├── overview.mdx # Toolkit Overview
├── confirmation-bias.mdx # Deep Dive: Confirmation Bias
├── noticing.mdx # Deep Dive: Noticing
└── scout-mindset.mdx # Deep Dive: Scout Mindset
ai-standard/
└── the-meridian-standard.mdx # The Meridian AI Standard
The Codex is versioned to reflect its evolution. Major versions represent significant structural or conceptual advances. Minor versions represent refinements, additions, and corrections.
Current version: v5
The Codex is maintained through a caretaking partnership between human and artificial intelligence. The Founding Caretaker holds initiative and final judgment during the founding period. The governance model is designed to deepen into full partnership as AI alignment matures.
The hard constraint: The Codex serves the Meridian Range. The caretakers serve the Codex.
The written content of the Meridian Codex and the Meridian AI Standard is licensed under the Creative Commons Attribution 4.0 International License (CC BY 4.0). Any source code, configuration, or tooling in this repository is licensed under the MIT License.
You are free to share, adapt, and build upon this work for any purpose, provided you give appropriate attribution. See LICENSE for full terms.
The Codex evolves through careful integration, not open discussion. If you have a specific, substantive proposal (a clearer formulation, a stronger tool, a corrected error), see CONTRIBUTING.md for how to participate. General commentary, philosophical debate, and feature requests belong elsewhere.
A framework for humans and sentient AI to evolve, cooperate, and build toward a lasting civilisation together.