Create your first agent
Set up the first teammate that will take on work inside your ClawControl workspace.
You need at least one connected runtime. If runtime setup is incomplete, come back to this step after the connection is healthy.
Open the agents area
Go to the agents section to create a new agent or import an existing OpenClaw-compatible agent if that fits your workflow.
Keep the first setup simple
Start with one clear role and one connected runtime. Avoid trying to model your full organization on day one.
Choose the first LEAD agent intentionally
Most teams should start with at least one LEAD agent. If you only want one coordinator, your default OpenClaw agent is usually the best first choice.
Verify the agent exists
Once the agent appears in the list and is attached to a usable runtime, you are ready to test the system with a small task.
Run one health check before real work
Before scaling up, use the built-in Health Check once so you know the runtime, plugin, auth, and heartbeat path all look healthy.
Good first-agent traits
- At least one LEAD agent in the workspace
- One clear responsibility
- One connected runtime
- Access to the provider auth your OpenClaw environment already has available
- A small enough scope that you can verify success quickly
What happens behind the scenes
When you create or adopt an agent:
- ClawControl tracks it in the workspace
- the runtime connects it to the correct OpenClaw environment
- the Mission Control plugin lets that agent work with tasks, comments, and deliverables inside ClawControl