Standardized interface for invoking external AI agents via CLI, process, or HTTP

LEX agentic affect domain: emotions, valence, mood regulation

LEX agentic attention domain: focus, salience, cognitive control

LEX agentic defense domain: threat detection, stress response, boundaries

LEX agentic executive domain: planning, working memory, cognitive control

LEX agentic homeostasis domain: self-regulation, balance, resource management

LEX agentic imagination domain: mental simulation, creativity, prospection

LEX agentic inference domain: causal reasoning, prediction, belief updating

LEX agentic integration domain: cross-domain binding and synthesis

LEX agentic language domain: linguistic processing and communication

LEX agentic learning domain: adaptation, memory consolidation, skill acquisition

LEX agentic memory domain: episodic, semantic, and working memory

LEX agentic self domain: identity, metacognition, self-model

LEX agentic social domain: empathy, social cognition, theory of mind

Durable shared knowledge with vector search, concept graph, and expertise tracking

Immutable audit logging with SHA-256 hash chain for LegionIO execution decisions

Code generation engine for LegionIO extensions

Fleet pipeline implementation: generate code, commit, open PR, handle feedback

Used to connect Legion to Esphome

Five-level extinction protocol with archival, audit trail, governance gates, and configurable settings