CORE

Constitutional rules create a deterministic operating system for personal AI assistants, establishing identity, personality, and response patterns that persist across every session. This ensures predictable behavior through self-contained core principles loaded at initialization.

Modularity breaks AI automation into discrete primitives—Skills, Commands, Agents, MCPs—that work together without tight coupling. This enables CLI-first workflows, progressive context loading, and code-before-prompts execution while maintaining clear governance and safety boundaries.

Constitutional rules provide a 3-tier context loading framework and role-based response style that apply safety and security protocols deterministically. This satisfies requirements for strict policy enforcement while enabling flexible automation across different workflows.

Load core principles and governance first, then layer modular components like Skills and Commands incrementally. This approach ensures identity and safety constraints are established before processing, enabling code-before-prompts workflows with strict history capture and infrastructure documentation.

Yes. Constitutional rules define infrastructure-agnostic principles for AI behavior that integrate with CLI-first tooling, structured documentation, and modular primitives. They work as a governance layer across different tool stacks and automation architectures.

Constitutional rules provide deterministic governance but require upfront definition of identity, personality, and policy constraints. Changes to core principles require session reinitialization; the framework prioritizes consistency and safety over real-time behavioral flexibility.