NXNX Assistant
NXNX Assistant is being built as a web-first personal AI workspace powered by the NXNX model stack.
The product direction is our own model, own agent architecture, own memory system, own workflow engine, and a local-first future with no dependency as core identity.
NXNX-S1 Local Model
Prototype
S1
User Mode
Clean assistant workspace
A simple public experience for planning, drafting, summaries, organization, and ideas. Sessions persist through the NXNX Assistant API when signed in.
Prototype Mode
Assistant
Offline fallback
Chat
Conversation is waiting for sign-in.
Quick Actions
Start faster
Tool System
Prototype
Run a prototype tool
Tool outputs are Prototype until NXNX-S1 is connected.
Workflow Library
Offline fallback
Saved workflows
Knowledge Library
Offline fallback
Personal knowledge
Memory
Memory overview
User memory context is empty.
Memory saves locally until sign-in.
Profile
User state
Profile uses offline fallback until sign-in.
Profile has not been saved this session.
Command workspace
Personal AI operating surface
A polished frontend foundation for chat, planning, drafting, memory, and workflows. Live generation is intentionally disabled until the NXNX model server is connected.
Prototype
Model Server Required
Chat workspace
Local Preview: I am wired as a frontend shell for the NXNX-owned model stack.
Plan my day, draft a launch note, remember project context, and prepare workflows.
Coming Soon: responses will route through agent-core, memory-engine, workflow-engine, and NXNX-S1.
Local Model Status
Model server offline
NXNX-S1 Local Model
Tool Router
Workflow launcher
Planner
Planning panel
Daily planning, weekly planning, project planning, goal tracking, task breakdown, and next actions.
Draft Studio
Drafting studio
Draft UI for email, documents, business writing, social posts, prompt creation, and notes.
Launch notePrototype
Editable generation will require NXNX-S1 and the response pipeline.
Memory Engine
Memory board
Saved preferences, important notes, personal context, and user facts for the future NXNX memory system.
Workflow Engine
Everyday workflow templates
Launchable surfaces for planning, drafting, search, and everyday work. Execution is not faked.
Developer Console
Test and evolve NXNX Assistant
Local Preview tools for testing prompts, memory, workflows, and future model-server contracts directly inside nxnx.one. No real model is connected yet.
Assistant Chat Tester
Conversation lab
Prototype
Local Preview
Model Offline
Response preview
NXNX-S1 model server is not connected yet. This is a local preview response.
System Prompt Editor
Local Preview
Active instruction
Current active prompt
Memory Tester
Offline fallback
Persistent memory scratchpad
Memory inspector is empty.
[]
Workflow Tester
Offline fallback
Execution debugger
{"name":"string","kind":"string","status":"string","input":{"description":"string","category":"string","steps":["string"],"favorite":"boolean"},"output":{"history":["execution"]}}
Select or run a workflow to inspect JSON.
Knowledge Inspector
Offline fallback
RAG-ready knowledge debugger
[]
{"chunks":[],"metadata":[],"attachments":[]}
Tool Inspector
Prototype
Tool execution debugger
[]
{"input":{},"output":{}}
Model Status Panel
Offline
NXNX-S1 Local Model
NXNX-S1 Control Panel
Prototype Mode
Model bridge
Context preview is waiting for session data.
No model bridge errors.
Improvement Notes
Local Preview
Testing notes
Future Backend Contract
Coming Soon
Model-server API
Returns server health.
Returns model name, model status, active checkpoint, context size, and memory status.
Accepts user message, system prompt, memory context, selected workflow, and conversation history.
Returns assistant response, metadata, response time, and model used.
Dataset and Training
Dataset/training status panel
Training artifacts are placeholders for the future NXNX-owned model pipeline.
Use `data/raw` for source material.
Instruction examples belong in `data/instructions`.
Planning, drafting, memory, and safety evals.
PyTorch/HuggingFace placeholder only.