The human layer for OpenClaw

The one-person company era begins.

Run agents, projects, and operations from one control surface built for real operators.

Install AgentOS Explore the system

Built on . Designed for real operators.

Mission Control Walkthrough

Why this lands

See the system

Workspaces, agents, runtimes, and health stay visible.

Dispatch real work

Missions map to actual folders, files, and operating policies.

Inspect what happened

Transcript history and created files keep output reviewable.

Control plane Sidebar + canvas + inspector

Deploy flow Plan -> scaffold -> run

Runtime truth OpenClaw-backed state

Operating style Local-first

Live topology canvas Mission dispatch Transcript-backed inspector Workspace bootstrap Planner deploy flows Onboarding and fallback

QuickStart

Install AgentOS and open Mission Control.

Pick an install path, run one command, and verify the local runtime.

# Fastest path. Install the published AgentOS release and open Mission Control locally.

curl -fsSL https://raw.githubusercontent.com/SapienXai/AgentOS-MissionControl/main/install.sh | bash

Packaged CLI install Launch with agentos start --open Verify with agentos doctor

# Best for repeat installs. Same control surface, cleaner upgrades through pnpm.

pnpm add -g @sapienx/agentos

Global package workflow Launch with agentos start --open Fast refreshes through pnpm

# Best for builders. Clone the repo, boot locally, and customize the control surface directly.

git clone https://github.com/SapienXai/AgentOS-MissionControl.git && cd AgentOS-MissionControl && pnpm install && pnpm dev

Ideal for forks and product teams Customize UI, flows, and branding Check openclaw gateway status --json

Start with the one-liner unless your setup already depends on pnpm or a source checkout.

What It Does

Everything the operator needs above the runtime.

This section mirrors the technical role of the reference site, but the cards are specific to AgentOS. These are the core operating surfaces already present in the product.

Live topology canvas

Visualize workspace, agent, runtime, and model relationships from one operator-facing map.

Mission dispatch

Route structured work to real OpenClaw agents with thinking levels and clear operational intent.

Transcript-backed inspection

Inspect runtime output, warnings, token usage, and created files after execution.

Workspace bootstrap wizard

Create a new workspace with templates, presets, memory, docs, and kickoff missions.

Planner deploy flows

Turn company context into a live workspace, agent team, automations, and first missions.

Agent presets and policies

Configure workers, browsers, setup agents, network rules, install scope, and heartbeat policies.

OpenClaw onboarding

Detect missing runtime state, install or start OpenClaw, and verify readiness directly from the UI.

Honest fallback mode

If the runtime is unavailable, the product says so clearly instead of pretending the system is live.

Works With Everything

Built to sit above the whole operating stack.

AgentOS does not replace the runtime. It rides above it. That means the control plane can unify models, files, sessions, automations, planner assets, and the real project workspace beneath.

OpenClaw GPT-5 Claude Local models Terminal Filesystem Git repos Workspaces Agents Runtimes Transcripts SSE snapshots OpenClaw GPT-5 Claude Local models Terminal Filesystem Git repos Workspaces Agents Runtimes Transcripts SSE snapshots

Planner Automations Gateway Policies AGENTS.md MEMORY.md Browser tasks Channels Team presets Docs Deliverables Local machine Planner Automations Gateway Policies AGENTS.md MEMORY.md Browser tasks Channels Team presets Docs Deliverables Local machine

Runtime truth Sessions, presence, gateway state, models, and transcript files

The live state stays grounded in OpenClaw instead of being duplicated into a fake SaaS abstraction.

Workspace reality Real project folders, docs, memory, deliverables, and scaffolds

Missions tie directly to the filesystem, which is where the work actually lives.

Operator actions Planning, deploy, inspection, onboarding, and policy-aware control

The human stays above the system with one coherent operating picture and real review loops.

Start with AgentOS

You are not just using AI. You are operating AI labor.

Run agents, projects, and operations from one control surface designed for real operators. Built on OpenClaw, AgentOS gives you structure, visibility, and control across workspaces, tasks, and execution.

Open repository Watch demo