Omni Core — Documentation
Everything you need to work in Omni Core — Nandeshou’s local-first AI workspace. Omni Core is a single desktop application that unifies a conversational assistant, a structured long-term memory called Kura, document libraries with search, and a context-aware document editor. It’s local-first by default — backed by a local Kura engine that starts with the app — with an optional Cloud mode you can switch on in Settings to sign in with your Nandeshou platform account and work against a hosted server instead. Start with Getting Started, then work through the guide for each area.
Omni Core also hosts a fiction-writing mode. If you write stories, see the dedicated Storyteller guide, which builds on the same workspaces, memory, and editor described here.
In this documentation
- Getting Started What Omni Core is, installing and launching the desktop app, creating your first workspace, and a tour of the window.
- The Workspace Omni Core's window layout: the menu rail, the docked panels and tab system, switching and saving workspaces, the Dashboard, and the Prompts panel.
- Kura Memory & Topics How Omni Core remembers: Kura's structured, typed topics with evolving memoranda, the topic tree, choosing which topics are active as context, and version history.
- Libraries & Search Document libraries in Omni Core: the default library, linking external folders, full-text search across your documents, and the four-view document editor with its document assistant.
- The Assistant Chatting with Kura in Omni Core: sending messages, approving and downvoting exchanges to shape memory, think levels and chat toggles, slash commands, and the in-editor document assistant.
- Models & Privacy Configuring Omni Core in Settings — the local-or-cloud backend and sign-in, licensing, LLM connections across local/OpenAI/Azure providers, model categories, and cost tracking — plus how the local-first design keeps your data on your machine by default.