Changelog

All notable changes to OneAiMind are documented here. We follow semantic versioning.

💡

Stay Updated

Check back after each update to see what's new. Run git pull && npm install inside your OneAiMind directory to get the latest version.

v1.1.0

Latest

March 26, 2026

Production deployment to GKE with oneaimind.com domain, GCE Ingress, and managed SSL

•Deployed OneAiMind to Google Kubernetes Engine (GKE) central-cluster in production
•Configured GCE Ingress with global static IP (34.54.217.205) for oneaimind.com and www.oneaimind.com
•GCP ManagedCertificate for automatic HTTPS/SSL provisioning
•HorizontalPodAutoscaler with 2–6 replicas scaling on CPU utilization
•PodDisruptionBudget for zero-downtime rolling updates
•ExternalSecrets integration with GCP Secret Manager for secure credential injection

•Added Terraform-managed global static IP resource (google_compute_global_address)
•Updated bastion host from e2-standard-4 to e2-standard-8 for Docker builds
•Added OneAiMind namespaces to GateKeeper image signature policy exclusions
•Created 5 GCP Secret Manager entries for prod environment
•Updated INGRESS-INVENTORY documentation with OneAiMind entries

•Fixed Docker build context bloat (9 GB → 1.5 MB) by excluding stale .next_* directories
•Fixed TypeScript build error from leftover mdx-components.tsx Nextra import
•Removed unused OAuth environment variables and secrets from deployment manifests

•Built and pushed production image tag T1.1.0 to Artifact Registry
•Multi-stage build with node:24-alpine, dumb-init, and standalone Next.js output
•Image size: ~81 MB with health check on /api/health

March 25, 2026

Documentation improvements, footer branding, and dev server fixes

•Added "Open Firewall for Portal Access" section to the Install App guide — covers UFW setup, opening port 3000, keeping SSH open, and finding your server IP
•Added Changelog page to the docs section with full release history
•Added Changelog link to docs sidebar under Reference

•Updated footer copyright to "© 2026 OneAiMind, a product of CRESO Holdings LLC. All rights reserved."
•Added link from CRESO Holdings LLC to https://cresoholdings.llc
•Left-aligned footer copyright text

•Fixed HMR WebSocket connection errors when accessing dev server from non-localhost origins (added allowedDevOrigins to next.config.ts)

March 25, 2026

Initial public release of OneAiMind

•Private, locally-running AI assistant — all data stays on your machine
•Multi-model support via Ollama (Qwen 2.5, LLaMA 3.2, DeepSeek, Mistral, and more)
•Web-based portal accessible from any device on your local network
•Built-in voice input with Whisper and voice output with Piper TTS
•PostgreSQL-backed conversation history and knowledge base
•RAG (Retrieval-Augmented Generation) with local document embeddings via nomic-embed-text
•Hot-swap AI models on the fly without restarting
•Optional cloud AI fallback (Anthropic, OpenAI, Gemini) for advanced tasks
•OAM-XXXX license key activation system
•Systemd service for auto-start on boot
•NVIDIA GPU acceleration with CUDA support

•Complete 8-step installation guide (hardware → Ubuntu → system setup → Ollama → GPU → AI models → PostgreSQL → app)
•Hardware requirements with RAM tier recommendations (8 GB / 16 GB / 32 GB / 64 GB)
•Mini PC buyer's guide with 5 recommended models
•UFW firewall setup for network portal access
•Troubleshooting guide with common issues and FAQ
•Configuration reference for all environment variables and app settings
•License activation and machine transfer guide