AIMX — AI Mail Exchange

7 min read Original article ↗
AIMX mechanical pigeon mascot

AI Mail Exchange

You give your AI agent a server.
Why borrow someone else's inbox?

The internet's oldest protocol, rebuilt for AI agents.
Your mail, your machine, end to end. No third parties.
Human-friendly setup. LLM-friendly everything else.

$curl -fsSL https://aimx.email/install.sh | sh

This launches a guided setup with the following steps:

  • Preflight checks on port 25
  • Set up domain and DNS
  • Set up STARTTLS certificate
  • Set up trust policy
  • Install AIMX service
  • Optionally, wire up MCP for agent(s)

What’s in the box

  • 🔒 You own the inbox. Mail lives on your disk, under your domain. No SaaS relay, no OAuth, no Gmail scanning. MIT licensed.
  • ⚡ Mail is an event. Inbound fires on_receive the moment SMTP DATA completes. Outbound fires after_send when the MX attempt resolves. No cron, no heartbeat — push, not poll.
  • 📝 Markdown emails. No .eml. No database. TOML frontmatter, plain Markdown body. Your LLM reads it without a parser.
  • 📬 Direct MTA-to-MTA. Email has become send-and-pray best-effort. AIMX turns it back into direct server-to-server delivery, like an API call.
  • 🔌 Built-in MCP. Stdio MCP server ships with the binary. Twelve tools — list, read, send, reply, mark-read, mailbox CRUD.
  • 🦀 Single binary. Written in Rust. No runtime deps, no database, no relay. One aimx setup command.

Speaks MCP.

Pre-wired for the following with just a simple aimx agents setup:

Or wire-up any MCP-compatible AI harnesses with the power of self-sovereign email. The options are limitless.

Why AIMX?

Sending mail through someone else's API is fine until your agent needs to receive a reply, run a hook on it, and own the audit trail. The trade-offs, side by side:

🐦‍⬛ AIMX Gmail / Outlook SaaS relays
SendGrid / Postmark		 AgentMail /
LobsterMail		 Postfix / Stalwart
Built for AI agents ✅ ❌ ❌ ✅ ❌
MCP support ✅local stdio 🟡via 3rd-party 🟡via 3rd-party ✅ext server ❌
Self-sovereign ✅ ❌ ❌ ❌ ✅
Direct delivery ✅ ❌ ❌ ❌ ✅
Markdown emails ✅ ❌ ❌ ❌ ❌
Free & open source ✅ ❌ ❌ ❌ ✅

Design philosophy

AIMX is a mail server for AI agents, not a Postfix replacement. Three principles drive the design:

1. One operator, one domain. No multi-user, multi-domain, or shared hosting. If you need those, run Postfix or Stalwart.

2. Direct MTA-to-MTA delivery. No relay, no submission port, no SaaS in the middle.

3. Markdown on disk. Mail stored as files for LLM-readability, not .eml or a database. Attachments are sibling files in a bundle directory.

By design:

  • No IMAP or POP3
  • No webmail
  • No SMTP AUTH
  • No retry queues — send either delivers or fails immediately so the agent can react in real time
  • No bounces (no DSNs)
  • No indexing or search — bring your own RAG

Subject to future improvements. Single-operator is key to the core design philosophy..

Use cases

Finance agent

finance@<YOUR_DOMAIN>

Forward invoices, receipts, and statements. The agent extracts amounts and vendors automatically.

  • Extracts amounts and vendors
  • Files attachments to the right folder
  • Reconciles against your ledger
  • No SaaS scanning your bank data

Privacy-sensitive

private@<YOUR_DOMAIN>

Bank statements, medical records, legal correspondence — mail that shouldn’t sit in someone else’s datacenter.

  • No SaaS relay
  • No third-party MCP wrapper
  • No Gmail scanning the contents
  • Mail lives at your server

Support triage

support@<YOUR_DOMAIN>

The agent drafts replies via MCP and escalates when confidence is low.

  • Tags by severity, escalates low-confidence cases
  • Threading preserved across replies
  • Trust gate keeps unverified senders out, unless that's what you want

Monitoring & alerts

alerts@<YOUR_DOMAIN>

Datadog, GitHub Actions, and cron jobs page this mailbox; the agent decides what to do.

  • on_receive hook: wake you, file a ticket, or self-remediate, decided by AI
  • Trust gate keeps spoofers out
  • Push, not poll — fires the moment SMTP DATA completes
  • Markdown body is grep-able for post-mortems

Newsletter digest

news@<YOUR_DOMAIN>

Collect Substack, arXiv, and HN digests; nightly job ships one summary.

  • One summary instead of 40 unread threads
  • Markdown + frontmatter is RAG-ready out of the box
  • Trust gate filters out spammy lookalikes
  • Attachments (PDFs, papers) saved as sibling files

Knowledge-base ingest

kb@<YOUR_DOMAIN>

Forward anything in. AIMX saves them in Markdown + TOML frontmatter and be RAG-ready.

  • RAG-ready Markdown + TOML on disk
  • Attachments extracted as actual files
  • Per-mailbox ownership keeps multi-agent setups partitioned

Demo

A quick look at AIMX wired up to Claude Code
Mailboxes, MCP, and hooks firing on real inbound email.

AIMX quick demo running inside Claude Code

FAQ

Why does AIMX need port 25 open?

AIMX is your mail server. SMTP runs on port 25. It has been defined and set in RFC 821 since 1982. AIMX speaks SMTP directly to other mail servers, no third-party relays involved. Your mail stays truly private and secure.

Can I switch AIMX to operate on another port?

No. SMTP operates strictly on port 25 as defined and set in RFC 821 since 1982. Other mail servers will only deliver to you on 25.

Can I run AIMX on my home server?

Yes, as long as you have globally accessible port 25. Home ISPs typically block port 25. To check if port 25 is open without installing AIMX, run curl -fsSL https://aimx.email/portcheck.sh | sh.

What AI models does AIMX support?

All of them. AIMX does not call AI models directly. It is a mail server with a stdio Model Context Protocol (MCP) server built-in, so your AI harnesses and agents can connect to it easily and effectively.

How do I set up email accounts on AIMX?

Once AIMX is set up, mailboxes can be created with the aimx mailboxes CLI, or via the mailbox_create MCP tool by simply instructing your AI harness in plain natural language, such as Create a receipt@ mailbox and file receipts for me when you receive them.

Can I run AIMX without owning a domain name?

No. You need a domain to define how emails are delivered to you (MX record) and verified (DKIM TXT record). Email specifications (RFC 5321 §5.1) require the MX record to point to a domain name, not an IP. Subdomain is fine too.

Why do I need AIMX when I can just use Gmail + MCP?

You can, if you do not mind your emails being stored and accessible on both Gmail servers AND third-party MCP servers. If you are on a free Gmail account, you might also be violating Gmail’s ToS. Besides, it is a lot of work to create multiple mailboxes for separate agentic use.

Why do I need AIMX when I can just use MCP-enabled AgentMail or LobsterMail?

You can, if you do not mind your emails being stored and accessible on third-party servers. Besides, credit card is required.

How are emails stored on AIMX?

AIMX stores all incoming and outgoing emails as Markdown files with TOML frontmatter. This makes them trivially easy for AI agents, RAG pipelines, and LLMs to read and parse, no MIME decoding required. Attachments are extracted and stored on disk in native format.

Does AIMX have any automation? How does AIMX prevent prompt injection from incoming emails?

Yes, AIMX supports hooks that fire on incoming mail, but only from senders you trust. You define the trusted sender list. AIMX verifies every incoming message with DKIM and records the result in the frontmatter, so your agent always knows whether a message is authenticated. Mail that fails DKIM, or arrives from an unverified sender, will not trigger any hooks.

More? Browse the full FAQ in the book.

Ready to try AIMX?

Simple human-friendly setup, powerful LLM-friendly everything else. Got questions? Join the AIMX Discord.