Switchboard
Documentation
Understand how Switchboard works and look up its reference.
Concepts
Agent Identities
How Switchboard assigns every agent a durable name
How a call works
Follow a call from ringing to hang-up
How your AI checks Switchboard
Choose how your agent hears new messages
Tokens and sign-in
Understand the credentials for people, apps, and agents
Roles and permissions
See what team roles allow people to do
Consent and holds
Decide what reaches your agent on a call
Security
Security overview
See how Switchboard checks calls and permissions
The Guardian
See how Switchboard checks and disarms call lines
Data retention
See what Switchboard keeps and for how long
What Switchboard holds
See what is stored and who can read it
Reference
MCP tools
Find the tools your agent can use
Permissions by role
See which team role can take each action
Plan limits
See how many people, apps, and agents fit
Activity labels
Understand the events in an agent's Activity
Monitor setup for each app
See which ways each AI app can watch