Interface Overview¶
Lumiverse's interface is built around a central chat view with a tabbed drawer that pulls in from the edge of the screen. Here's a tour of what's where.
Main Layout¶
+--------------------------------------+----+
| | |
| Chat Area | T |
| | a |
| | b |
| | s |
+--------------------------------------+ |
| Input Area | |
+--------------------------------------+----+
- Chat Area — Where messages appear. Shows the conversation with your character, including avatars, expressions, and (optionally) a per-chat wallpaper background.
- Input Area — Where you type messages. Includes action buttons for attachments, persona switching, quick replies, add-on toggles, and dry runs.
- Drawer — A single docked drawer that hosts every panel as a tab. Pinned tabs appear on the visible edge so you can switch with one click; the rest live behind an overflow menu.
- Chat Heads (optional) — Floating circular avatars that follow the screen edge and act as quick-switchers between recent chats.
Drawer panels can be docked to the opposite edge using the Spindle dockPanels system, and on mobile the drawer slides in as a sheet.
Drawer Tabs¶
The drawer hosts every workspace panel as a tab. You can reorder them with drag-and-drop and hide ones you don't use from the tab-bar context menu.
Character & Story¶
| Tab | Purpose |
|---|---|
| Profile | View and edit the active character's card |
| Characters | Browse, search, import, and manage your character library |
| Personas | Create, switch, and manage user personas |
| Lorebook | Edit world books and lorebook entries |
| Pack Browser | Browse and install content packs (Lumias, Looms, themes, tools) |
| Creator Workshop | Author your own Lumia items and Loom presets |
Generation Configuration¶
| Tab | Purpose |
|---|---|
| Reasoning | Configure chain-of-thought, reasoning effort, prompt prefix/suffix, and start-reply-with |
| Loom | Configure narrative structure, story beats, pacing, Sovereign Hand, and director cues |
| Composition | Pick which Lumia/Loom content is active, set context filters, and tune prompt assembly |
| Connections | Manage LLM, Image, Speech-to-Text, and Text-to-Speech API connections |
| Council | Configure the Lumia Council, tool functions, and sidecar agents |
| Summary | Configure context summarization and truncation |
Memory & Knowledge¶
| Tab | Purpose |
|---|---|
| Memory Cortex | Browse entities, relationships, font colors, and stats. Manage vaults and interlinks here. |
| Databank | Upload reference documents that the AI can pull from during generation |
| World Info | Live readout of which lorebook entries activated for the current generation |
| Council Feedback | Inspect the latest council execution — what the tools did and why |
Visual & Output¶
| Tab | Purpose |
|---|---|
| Image Generation | Configure and trigger AI image / scene generation |
| Wallpaper | Set global or per-chat backgrounds (images, video, animated GIFs) |
| Theme | Customize colors, accents, fonts, glass effects, and CSS overrides |
| Regex Scripts | Author find/replace transformations for prompts or rendered output |
Meta¶
| Tab | Purpose |
|---|---|
| OOC | Out-of-character comment display settings |
| Branch Tree | Visualize and navigate the chat's branch history |
| Dream Weaver | Generate brand-new characters from prompts |
| Extensions | Install and manage Spindle extensions |
Spindle extensions can register additional drawer tabs that appear alongside the built-ins.
Chat Controls¶
Inside an active chat, each message exposes:
- Regenerate — Re-roll the last AI response (creates a new swipe)
- Continue — Ask the AI to continue writing from where it left off
- Swipe arrows — Navigate between alternative responses on the same message
- Edit — Click on any message to edit its content in place
- Branch — Fork the conversation at any message into a separate timeline
- Author's Note — Inject a system-level instruction at a configurable depth
Input Area Actions¶
The input area exposes several actions beyond just sending messages:
- Attachments — Upload images, audio, or documents to include with your message
- Persona switcher — Quickly change which persona is active for this chat
- Quick Replies — Insert pre-written response templates
- Guided Generation — Enable structured output guidance
- Add-ons (Puzzle icon) — Toggle persona add-on blocks on and off, including global add-ons attached to the active persona
- Dry Run — Preview the exact prompt the AI will see without sending a real request
- Voice input — Dictate via the configured Speech-to-Text connection
Per-chat toggle state for add-ons is remembered, so flipping a block off in one chat doesn't affect another.
Command Palette¶
Press Cmd+K (macOS) or Ctrl+K (Windows / Linux) to open the Command Palette. This is the fastest way to navigate Lumiverse and gives quick access to:
- Drawer tabs and modal panels
- Settings sections
- Chat-specific actions (regenerate, swipe, branch, etc.)
- Extension-registered commands
Type to search, then press Enter to execute.
Landing Page¶
When you open Lumiverse (or navigate to the home page), you land on the Landing Page. It shows recent chats grouped by character. Click any chat to resume it, or click a character to start a new conversation. The landing page layout (compact list vs. grid) and the number of chats shown are configurable from Settings → Display.
Settings¶
Click the gear icon (or open the Command Palette and search "Settings") to open the Settings modal. The sidebar groups settings into categories:
Account & Display¶
| Section | What's Inside |
|---|---|
| Account | Username, password, avatar |
| Display | Modal sizing, pagination, toast positions, landing layout, chat heads |
| Chat | Message-per-page, enter-to-send, draft saving, message render options |
| Notifications | Push notification preferences |
AI & Generation¶
| Section | What's Inside |
|---|---|
| Embeddings | Vector embedding provider and indexing config |
| Memory Cortex | Entity tracking, salience, sidecar, memory cortex config |
| Web Search | SearXNG integration for the web-search tool |
| Tokenizers | Manage tokenizer downloads for accurate token counting |
| Voice | Speech-to-Text and Text-to-Speech defaults |
| MCP Servers | Manage Model Context Protocol server connections |
| Guided Generation | Default templates for guided output |
| Quick Replies | Manage the input-area quick-reply library |
Extensions¶
| Section | What's Inside |
|---|---|
| Extensions | Manage installed Spindle extensions |
| Extension Pools | Per-extension storage quota overrides |
Operations¶
| Section | What's Inside |
|---|---|
| LumiHub | LumiHub account and content sharing |
| Users | (Multi-user installs) Manage additional user accounts |
| Data Portability | Export / import everything, API keys, tickets |
| Migration | Import from SillyTavern (Local, SFTP, SMB, Google Drive, Dropbox); re-apply TagLibrary backups |
| Operator Panel | Check for updates, switch branches, restart server (requires the runner) |
| Advanced | Power-user toggles |
| Diagnostics | System health, version, embedding/cortex status |
Mobile
On smaller screens, the drawer becomes a slide-in sheet. Swipe from the edge, tap any pinned tab icon, or use the Command Palette to navigate. Lumiverse also supports PWA installs — add it to your home screen for an app-like experience.