Comment by kellpossible2

what's to stop an agent creating an outbound call with the var to a malicious endpoint? (unless you whitelist what it has access to)