Once an A2A Server exists, its header and Settings are where you put it live for other agents to call, see which agent it fronts, turn on streaming, and manage how callers authenticate.

Overview

The header shows the server's name, description, an Active/Inactive toggle, when it was last updated, a Manage API button, a Deploy (or Undeploy) button, and a ⋯ menu with Edit and Delete. The side navigation starts with an Agent Exposed card and switches between Skills and Settings.

Deploying or Pausing the Server

Click Deploy in the header — the server goes Active (shown as Live in the list) and starts answering A2A requests. The same button becomes Undeploy to pause it; the small header toggle flips the same state.

Note: New servers start paused.

Agent Exposed

The card at the top of the side navigation, labeled Agent Exposed, shows the agent this server fronts (avatar and name). Clicking it opens that agent's own configuration — handy for jumping straight to the behavior your callers actually reach.

Settings

Headed Settings — "Configure advanced features for your A2A server" — with one option: a Streaming toggle, "Enable streaming of responses between agents." Turning it on advertises streaming on the server's agent card so callers can use the streaming endpoint (/jsonrpc/sse); it auto-saves.

Authentication

Caller access is set up in API Manager. Click Manage API in the header to open the server's API group, then give callers access through API clients and the access profiles that are granted this server's API group.

Deleting the Server

Header ⋯ menu › Delete. A confirmation dialog names the server; deleting cannot be undone, and you're returned to the A2A Servers list.