neuromancerOS Explore the product
THE CONTROL LAYER FOR AI WORK

One project.
Every agent.
Any screen.

AI coding agents are powerful. The hard part is keeping the right project, computer, context, and conversation together. Neuromancer OS gives all of that work one place to live, from desktop and phone to smart glasses.

See how it works See what’s built
Smart glasses clients for Even Realities G2 and Meta Display.
N NEUROMANCER / WORKSPACE
PREVIEW
PROJECT WORKSPACE 01 / 04
Your work, in focus.

Choose the project and where the work should happen.

YOUR NEXT MOVE ILLUSTRATIVE PREVIEW
“Open my project and show me what I was working on.”
Project selected · ready for the next step↗
Illustrative product flow. No live accounts or devices are connected to this page.
PROJECT-AWARE WORKNATIVE AGENTSVOICE CONTROLPRIVATE CONTEXTLOCAL MODELSHUMAN APPROVAL
01 / THE IDEAONE LAYER ABOVE THE AGENTS

More agents won’t fix a broken workflow.

Today, your projects sit in one place, your AI conversations in another, and your computers somewhere else. Every time you switch screens or tools, you rebuild the context by hand.

Neuromancer OS brings those pieces into a single project-first workspace. Pick what you’re working on, choose the agent and computer that should handle it, and keep the human decisions visible.

01 —
◈

Start with the project

A private project library holds the source material. Find relevant context before you open or continue an agent conversation.

02 —
⌘

Choose where it runs

Work through native coding agents on a selected computer, or use a local model when that’s the right fit. The destination stays explicit.

03 —
◎

Stay in the loop

Speak or type, review before sending, handle approvals, and hear the useful result. The agent does the work; you keep the judgment.

02 / THE FLOW

From “I have an idea”
to “show me the work.”

One clear path through the moving parts of agent work. No guessing which machine, project, or conversation is active.

  1. 01

    Name the project

    Start with the work itself and the context that belongs to it.

    ↗
  2. 02

    Pick a destination

    Choose the computer, available agent, and model deliberately.

    ↗
  3. 03

    Give the instruction

    Speak or type. Review the request before anything is sent.

    ↗
  4. 04

    Follow the result

    See the actual conversation and approvals, then continue from there.

    ↗
03 / THE INTERFACESWORK WHERE YOU ARE

The work travels with you.

One idea, expressed through the screen you have. The experience adapts; the project and destination stay clear.

DESKTOP

Full view. Full control.

See projects, agent sessions, status, and approvals together in a workspace built for real work.

PHONE

Continue from anywhere.

An installable web app makes project selection, conversations, voice, and review available in your hand.

SMART GLASSES · EXPERIMENTAL

Less screen. More flow.

Clients for Even Realities G2 and Meta Display bring project and agent navigation to smart glasses. Full physical, hands-free validation is ongoing.

04 / WHERE WE AREAN EARLY WORKING PRODUCT

Already real.
Still growing.

Neuromancer OS is being built in stages. The core workspace exists today; the deeper cross-device and hands-free experience is the next chapter.

BUILT
  • Web and phone workspace
  • Smart glasses clients for Even Realities G2 and Meta Display
  • Authenticated computer hubs
  • Native agent and local-model paths
  • Private project library
  • Voice navigation and explicit approvals
  • Durable tasks across devices
  • Approved, sourced project memory
  • Safer private-context coding
  • Fully validated hands-free workflows
THE NEXT INTERFACE FOR WORK

The agents are here.
Now give them a place to work.

Neuromancer OS is an independent project exploring how people can direct many AI agents without losing the context, control, or continuity that makes the work matter.

Explore the preview