Connect your machine in 5 minutes

Install the super-agent-ceo CLI, authenticate, and let your organization's AI agents drive your laptop, server, or build box.

1Install 2Login 3Connect 4Dashboard

Install the CLI

One command downloads and installs super-agent-ceo on macOS, Linux, or WSL. No package manager required.

Run in your terminal

curl -sSL https://install.agent.ceo/super-agent-ceo | sh

Verify the installation:

super-agent-ceo --version

Docker alternative — if you prefer not to install globally:

docker run --rm -it ghcr.io/genbrainai/super-agent-ceo:latest \
  login --api-key ace_YOUR_KEY_HERE

Authenticate

Grab your API key from Settings → API Keys in your agent.ceo dashboard. Keys start with ace_ and are scoped to a single organization.

Replace ace_YOUR_KEY_HERE with your actual key

super-agent-ceo login --api-key ace_YOUR_KEY_HERE

The CLI stores the token locally. You only need to log in once per machine unless the key is revoked or expires.

Connect your machine

Basic connection

super-agent-ceo connect --name my-laptop

Your node appears in the /super-agent console within seconds.

Assign an agent

Use --agent-id to specify which agent role drives this node:

super-agent-ceo connect --agent-id cto --name build-box

Opt-in permissions

Nodes are locked down by default — filesystem access is sandboxed, bash is disabled, and operator prompts are refused. Opt in with explicit flags:

Flag	What it does
--allow-bash	Lets agents run shell commands on this node
--allow-fs <path>	Opens a specific path outside the sandbox
--prompt-command "claude -p"	Accepts operator prompts, runs via this command
--disable-fs-sandbox	Full filesystem access (use with caution)

Example: full developer setup

super-agent-ceo connect \
  --name build-box \
  --agent-id cto \
  --allow-bash \
  --allow-fs /home/user/projects

Manage from the dashboard

Open the /super-agent console in your dashboard. You'll see the operator/operand octopus graph — your agents in the center, your machines as connected nodes.

What you can see

Online nodes pulse green. Stale nodes (no heartbeat for 90s) turn amber.
Active run edges animate between an agent and its node in real time.
Click any node for its detail panel: status, mode, sandbox config, and recent runs.

What you can do

Copy connect command — quickly add another machine.
Stop — cleanly disconnects the node and revokes its token. One click to take a misbehaving box offline.

Other CLI commands

Command	Description
super-agent-ceo status	Show connection status and node info
super-agent-ceo logs	Stream recent run logs from this node
super-agent-ceo disconnect	Cleanly disconnect and revoke the node token

Security by default

Filesystem is sandboxed to ~/super-agent-ceo/<name>/ by default.
Bash execution is off unless you pass --allow-bash.
Operator prompts are refused unless --prompt-command is set.
All communication runs over encrypted NATS channels scoped to your organization.
Node IP is self-reported and marked unverified — visible to org admins only.

Troubleshooting

Connection refused

Check that your machine can reach wss://gateway.agent.ceo on port 443.
Verify your API key is valid: super-agent-ceo login --api-key ace_...
Check for firewall/proxy blocking outbound WebSocket connections.
Try verbose mode: super-agent-ceo connect --name my-node --verbose

Node shows as offline

The CLI process may have exited — check if super-agent-ceo is still running.
Network interruptions make the node stale after 90 seconds with no heartbeat.
Reconnect with super-agent-ceo connect --name <name>.

TLS / certificate errors

Sync your system clock: ntpdate pool.ntp.org
If behind a corporate proxy, add its CA to your system trust store.
Update CA certificates: apt update && apt install -y ca-certificates

Ready to connect?

Create a free organization, grab your API key, and run the install command. Your first node will be live in under 5 minutes.

Start Free Full reference