Design Principles

The following Design Principles form the architectural foundation of POOKA. They describe how information should be organized to enable sustainable collaboration between humans and AI. They are normative principles that guide the architecture itself rather than specific technical implementations.

Source of truth

Each principle below carries its canonical title and its opening statement verbatim from Chapter 7, which remains authoritative. Follow the link under a principle for its full rationale.

Meaning Before Mechanics

Architecture begins with meaning, not implementation.

Section 7.1 in the Design Paper

Context Before Prompt

Context is an architectural concept rather than a conversational construct.

Section 7.2 in the Design Paper

Explicit Semantics

Semantics should never rely solely on interpretation.

Section 7.3 in the Design Paper

Explicit Relationships

Knowledge does not exist in isolation.

Section 7.4 in the Design Paper

Boundaries Are Architecture

Boundaries define the architecture as much as the information itself.

Section 7.5 in the Design Paper

Representation Must Be Explicit

An Actor never implicitly represents an Identity.

Section 7.6 in the Design Paper

Structure Before Automation

POOKA assumes that Artificial Intelligence operates most effectively on well-structured information.

Section 7.7 in the Design Paper

Minimum Necessary Context

AI should only receive the information required to perform the intended task.

Section 7.8 in the Design Paper

Local First, Cloud When Necessary

Architectural decisions should preserve autonomy wherever practical.

Section 7.9 in the Design Paper

Architectures Outlive Implementations

Architectural concepts should survive technological change.

Section 7.10 in the Design Paper

From principle to implementation

The Design Principles guide the architecture. Chapter 11 describes what a conforming implementation should preserve, and section 11.8 defines conformance: an implementation conforms to POOKA when it preserves the Core Concepts, their Relations and their Boundaries, maintains the conceptual distinction between the Knowledge, Governance and Behavior layers, and remains consistent with these Design Principles.

Conformance is a matter of architectural fidelity rather than certification. The paper defines no compliance tests, certification criteria or mandatory technologies.