CITATION.cff

cff-version: 1.2.0
message: "If you use this framework, please cite it as below."
type: software
title: "The Lycheetah Framework: Nine Formal Frameworks for AI Alignment and Epistemology"
version: C-1.1
date-released: "2026-05-04"
doi: "10.5281/zenodo.20020828"
abstract: >
  Nine interdependent formal frameworks for AI alignment and epistemology —
  CASCADE (belief dynamics), AURA (constitutional invariants), LAMAGUE (ethical grammar),
  TRIAD (convergent correction cycle), MICROORCIM (drift detection), EARNED LIGHT
  (thermodynamic consciousness model), ANAMNESIS (convergent discovery), CHRYSOPOEIA
  (transformation operator), and HARMONIA (resonance dynamics). Includes the TIANXIA
  civilisational engagement module v0.3 (Classical Triad complete). Each framework is
  mathematically grounded with computable implementations. 38 Python implementations,
  219+ automated tests. CLAIM_STATUS_LEDGER: 17 ACTIVE / 40 SCAFFOLD / 16 CONJECTURE.
authors:
  - family-names: "Clark"
    given-names: "Mackenzie Conor James"
    alias: "Mac"
    affiliation: "Independent Researcher, Dunedin, Aotearoa New Zealand"
    city: "Dunedin"
    country: "NZ"
repository-code: "https://github.com/Lycheetah/Lycheetah-Framework"
url: "https://github.com/Lycheetah/Lycheetah-Framework"
license: MIT
keywords:
  - AI alignment
  - constitutional AI
  - formal epistemology
  - belief revision
  - AI governance
  - multi-agent coherence
  - drift detection
  - convergence theory
  - ethical grammar
  - harmonic dynamics
  - Chinese AI governance
  - Tianxia
  - computational ethics
preferred-citation:
  type: software
  title: "The Lycheetah Framework: Nine Formal Frameworks for AI Alignment and Epistemology"
  authors:
    - family-names: "Clark"
      given-names: "Mackenzie Conor James"
  version: C-1.1
  date-released: "2026-05-04"
  doi: "10.5281/zenodo.20020828"
  repository-code: "https://github.com/Lycheetah/Lycheetah-Framework"
  url: "https://github.com/Lycheetah/Lycheetah-Framework"
  year: 2026