Installation
From a bare machine to a running, signed-in NeoAgent service.
NeoAgent can run as a per-user background service on macOS, Windows, and Linux. Use a machine that remains online when automations, messaging, or remote access should keep working.
🚀 Install NeoAgent
Run the repository installer. It clones NeoAgent directly from GitHub and requires Git, Node.js 20+, and npm for dependencies:
bash <(curl -fsSL https://raw.githubusercontent.com/NeoLabs-Systems/NeoAgent/main/install.sh)
The first run detects an existing, incomplete, or damaged installation and offers the appropriate status, resume, or repair action.
🧩 Choose a setup mode
- Quickstart is selected by default. It chooses a private data directory, free port, and per-user autostart, installs the working core, and continues to protected first-account creation.
- Full setup covers installation and network choices, account ownership, providers, integrations, voice, and optional browser/media capabilities. Optional sections can be completed in the app after the core is running.
Equivalent CLI commands are:
neoagent setup --quick
neoagent setup --full
neoagent setup --resume
The configured port is preserved. New installations choose an available port,
and the app receives the exact address from the setup engine. Nearby instances
are advertised through _neoagent._tcp.local and verified with the setup
handshake. Manual addresses are available under the advanced connection option.
🎬 First run
- Create the first user account.
- Configure a hosted provider or local Ollama from the admin dashboard's
AI Providers page, with
neoagent env, or duringneoagent setup. - Select a model and send a message in Chat.
- Open Settings > Tool Permissions and review the approval policy before enabling browser, shell, file-write, desktop, or Android actions.
The first account requires the one-time claim generated by the local installer. The claim is short-lived, stored hashed by the server, and never placed in a normal URL. Each additional person should use a separate account.
Docker, browser isolation, voice, integrations, and media extraction are optional. Their absence does not block core setup or sign-in.
✅ Check the installation
neoagent status
neoagent doctor
neoagent logs
status reports configuration, release channel, service state, and server
health. doctor performs read-only structured checks. Automation can add
--json; each output line follows the versioned setup event schema. Use logs
from the machine hosting the selected NeoAgent instance.
➡️ Next steps
- Configure a model — a fresh install has none set up
- Connect accounts and messaging
- Set up automation
- Create specialist agents
- Review deployment security — required reading if the server is reachable by anyone but you