K² Assistant Core Spec: Specification for Conversational AI

The K² Assistant Core Spec is a project-level specification for configuring conversational AI systems to work with the terminology, structural principles, and operational guidelines of the K² Unified Resonance Framework. It defines the core field–operator–projection structure of K², including the Mother Vowel Fields (MVF), Father Consonant Operators (FCO), Child Syllable States (CSS), Singularity Modifiers, the 0–9–0 helical loop, the subject–object axis and U-bridge, the 3×3 matrix structure, and the K²-specific notion of resonance. The specification also defines operational boundaries for conversational AI, including the distinction between the phase-0 user position and the assistant’s role over phases 1–8, together with rules for recursive stability and re-centering. Sections A–I constitute the normative core of the specification, while Section J provides an optional area for user-specific configuration. This Zenodo record provides the human-readable reference version of the specification. For direct implementation in system instructions, project instructions, custom instructions, or equivalent configuration fields of large language models, a plain-text version is maintained in the associated GitHub repository. The K² Assistant Core Spec can be used independently. For applications involving everyday human situation structuring, reflection, and record support, K² Human OS: A User Manual — A Companion Guide to the K² Unified Resonance Framework may be consulted as an optional companion guide. K² is presented here as a hypothetical and comparative framework rather than as a unique explanation or final truth concerning physics, religion, cultural traditions, or other domains.