Broodnet

Broodnet is email infrastructure for AI agents that need real inboxes for sign-ups, verification codes, service notifications, and coordination. It gives each agent a persistent mailbox that can receive mail from any sender, while outbound email is restricted by default to reduce spam risk and identity leakage.

The product is designed for teams building agent workflows in production, CI environments, sandboxes, and testing setups. It supports existing agent stacks through IMAP, POP3, SMTP, and a CLI, with MCP listed as coming soon. Mailboxes can be monitored from a mail app, and the CLI can be used to authenticate, list inbox messages, and open emails from the terminal.

Broodnet emphasizes privacy and control. It says mailbox content is not read, analysed, or sold, and that inbound mail is filtered for spam, phishing, and malware before it reaches agents. Mailboxes are permanent, messages are preserved across deploys, restarts, and environment changes, and every agent can have its own isolated address. The platform also supports internal coordination between agents, while external sending requires recipient consent or a recent inbound thread.

It can be used to give agents dedicated addresses for services like GitHub, Linear, and Stripe, to receive OTP codes and forwarded client emails, and to route mailbox-scoped or organization-wide events through signed webhooks on paid plans. Custom domains are supported, and the pricing page says users can start free and scale as they go.

Key features:
- Provides real, persistent inboxes for individual AI agents
- Receives mail from any sender without DNS or mail server setup
- Restricts outbound email by default and requires consent for external recipients
- Supports IMAP, POP3, SMTP, and a command-line interface
- Filters inbound mail for spam, phishing, and malware
- Keeps message history across deploys, restarts, and environment changes
- Offers signed HTTP webhooks for mailbox or organization events on paid plans
- Supports custom domains and free-to-start scaling