Elia: governed hybrid neuro-symbolic architecture where neural intelligence is a capability, never the authority.

Symbolic governance always in charge — LLMs are optional & validated.

Key features: strict separation of concerns, graceful degradation, audit trails, circuit breakers, heartbeats, SLOs, anti-deadlock.

Early stage (detailed spec in EL-ARCH.md). Phase 0 (minimal skeleton) coming soon. Not a full-time coder — open to implementation help.

Seeking feedback from systems architects, AI safety people, engineers: - governance model viability - biggest risks / blind spots - good first prototype domain (medical, monitoring, assistants, agents, education, ... any safety-critical domain)

Repo: https://github.com/Jmc-arch/elia-governed-hybrid-architectur...

Thanks for any thoughts! Jmc