Set up your Escattle system
A short walkthrough that wires up your local AI agents framework. Takes about three minutes; you can revisit any step before deploy.
Name your system
A short, lowercase identifier used in service names, paths, and the dashboard header. You can change the display name later.
Create the admin account
This account signs in to the dashboard and approves new agents. You can add more accounts after setup.
Connect to Anthropic
Pick how this system authenticates. Whatever you choose stays on this machine and never leaves your infrastructure.
Use one method only — don't set both.
How will you reach the dashboard?
Choose where Escattle accepts connections. You can change this later by editing your tenant config.
Create your first agent
Each agent is a long-lived Claude Code subprocess with its own role and memory. Start with one; add more from the dashboard.
Wire in optional integrations
All off by default. Anything you skip stays available — turn it on later from the dashboard.
Review and deploy
Everything below will be written to your tenant configuration and used to bootstrap the system.
System running
Your Escattle system is live. The dashboard is reachable at the URL below; the CLI commands let you operate it from a terminal.