Give your agent a safe inbox: prompt-injection screening, suppression lists, and idempotent sends

An agent with an inbox has three ways to hurt itself. It reads a stranger's email and follows the instructions inside it. It keeps mailing an address that bounced or complained, until the sending repu