Complete documentation of: - Read flow (vault browser) - Search flow (embeddings + pgvector) - Edit flow (4-step GRM workflow with branch → MR → review → merge → auto-sync) - Agent context flow (real-time execution logs) - System architecture (K8s topology) - Pod infrastructure (5-6 core + 6-8 supporting pods) - Traffic flow diagram - Deployment checklist Total pods documented: - 2× poimen-memory (API server, HA) - 2× memory-db (PostgreSQL primary + replica, 20Gi each) - 1-2× frontend (React SPA, vault UI) - 5+ ArgoCD pods (CD orchestration) - 1× nginx-ingress (reverse proxy)
