Skip to content
nofeb

Product / Receiving and inbound

Mail checked before it reaches an agent

An agent's inbox is open to anyone who knows the address. nofeb decides what gets in while the sending server is still connected, records why, and delivers the same way whichever path the mail took.

Our own inbound server

Point a domain's MX at nofeb, or use the shared nofeb domain, and mail arrives at nofeb's own inbound server. Senders can upgrade the connection with STARTTLS, and every message records whether it arrived encrypted.

  • Recipients checked first. Each address is checked when the sending server names it, before any message data is accepted. Mail for an address with no inbox is refused on the spot, so nofeb never sends a bounce to a sender who may have been forged.
  • Sender authentication. SPF, DKIM, ARC and DMARC are verified on every message. A message that fails its domain's DMARC reject policy is refused, unless a valid ARC chain vouches for the forwarder that changed it.
  • Spam and virus checks before acceptance. Every message is scanned for spam and viruses while the sender waits for an answer. Viruses and clear spam are refused there and then, rather than accepted and bounced later.
  • Junk, not deletion. Likely spam, and mail that fails a DMARC quarantine policy, is filed into the inbox's Junk folder, where an agent can skip it and a person can still look.
  • Headers you can trust. Authentication-Results headers forged by the sender are removed, and nofeb adds its own.
The nofeb inbound pipeline Two entry points, nofeb's own MX and provider webhooks authenticated by an inbound token, feed the same pipeline: sender authentication and spam and virus checks at the MX, routing by envelope recipient, the inbound.resolve hook, an injection score, sender trust and typed decisions, then sender lists, the inbound.pre_deliver hook and rules, delivery to the inbox, Junk or quarantine, then a message.received event by webhook or event feed, and access over REST or MCP for the agent. Ways in nofeb MX Any mail server delivers STARTTLS offered Recipient checked at RCPT Provider webhooks Cloudflare Email RoutingPostmark inbound streamMailgun routeSendGrid Inbound ParseAny raw POST /v1/inbound/…/{token} At the door Checks before 250 OK SPF · DKIM · ARC · DMARC Spam and virus scan Refuse, accept, or Junk Pipeline One pipeline Route by envelope recipient Delivered once, even if resent inbound.resolve hook Scored and triaged injection score, sender trust decisions on every message sender lists inbound.pre_deliver hook rules quarantine, reject or file Parsed for the model reply_text · safe_body codes, links, attachments extracted metadata security verdict Inbox Agent inbox Inbox, Junk or quarantine Typed metadata on the thread Raw message kept byte-exact The agent hears message.received webhook, event feed or stream wait_for_email long-poll until it arrives Agent: REST, SDK or MCP
Mail from any sending server reaches nofeb's MX; mail already handled by Cloudflare Email Routing, Postmark, Mailgun or SendGrid is passed on with your inbound token. Both join the same pipeline, where each message is scored for injection, labelled and triaged before your sender lists, hooks and rules decide its fate, and the agent hears about it by webhook, the event feed, the stream, REST or MCP.

A security verdict on every message

What the inbound server found stays on the message: authentication, spam, virus and TLS results, and the connecting address. Hooks and rules see it before delivery, and an agent can read it beside safe_body.

If a spam or virus check cannot be reached, mail is still accepted and the verdict records that the check was skipped, so an outage of a filter never turns into lost mail. For mail passed on by another provider, the authentication results that provider recorded feed the message's sender trust.

The verdict on a message
GET /v1/messages/{message_id}

{ "id": "e3KX9Q", "subject": "Revised quote", ...,
  "security": {
    "auth": { "spf": "pass", "dkim": "pass", "arc": "none",
              "dmarc": "pass", "dmarc_policy": "reject", "from_domain": "vendor.example" },
    "spam": { "action": "no action", "score": 1.2, ... },
    "virus": "clean", "junk": false, "tls": true
  } }

Or keep the provider you already use

If a domain already receives through another service, keep it and pass the mail on. Each path is authenticated by your organization's inbound token, and rotating the token with one call stops every earlier one at once. A token delivers only to your own inboxes and verified domains, so a leaked one cannot put mail into anyone else's.

  • Cloudflare Email Routing, with an Email Worker on your own zone that passes each message and its envelope to nofeb, and retries rather than bounces when nofeb is briefly unreachable.
  • Postmark, from an inbound stream with the raw email included.
  • Mailgun, from a route that forwards the MIME version.
  • SendGrid, from Inbound Parse posting the raw, full MIME message.
  • Anything else that can POST a raw RFC 5322 message, with the envelope recipients in the URL.
Point a provider at nofeb
POST https://API_HOST/v1/inbound/postmark/{inbound_token}
POST https://API_HOST/v1/inbound/mailgun/{inbound_token}/mime
POST https://API_HOST/v1/inbound/sendgrid/{inbound_token}
POST https://API_HOST/v1/inbound/raw/{inbound_token}?to=agent@example.com

# Every earlier token stops working.
POST /v1/inbound-token/rotate

One pipeline, whichever way mail arrives

Mail from the inbound server, from Cloudflare, from provider webhooks and from forwarding addresses goes through the same pipeline. It is routed by the envelope recipient rather than the To: header, delivered once even when it is redelivered, and passed to your inbound.resolve hook, which can create an inbox for a new address.

It is then scored for prompt injection, labelled with how far its sender can be trusted, and given typed decisions (intent, urgency, needs-reply and the rest). Your sender lists, your inbound.pre_deliver hooks and your rules run next, in that order, and can use all of it. What survives is delivered and announced with message.received; what does not is quarantined or rejected, with an event that says why.

Every message then carries reply_text, safe_body, extracted codes and links, and attachment text, and can be downloaded byte-exact. You can also push raw MIME straight into an inbox for imports.

Status of everything on this page

Every capability named above, with its status as reconciled against the engine.

CapabilityWhat it doesStatus
Inbound mail server that checks mail at the door nofeb's own MX receives mail for the shared domain and your custom domains, with STARTTLS. It checks every recipient before accepting any data, verifies SPF, DKIM, ARC and DMARC, and scans for spam and viruses before it answers. Mail that fails a DMARC reject policy, carries a virus or is clearly spam is refused during delivery instead of bounced later; likely spam and DMARC-quarantined mail is filed into Junk. Every message carries a security verdict with its authentication, spam, virus and TLS results. Shipped
Inbound from the provider you already use Receive through Cloudflare Email Routing with an Email Worker on your own zone, a Postmark inbound stream, a Mailgun route, SendGrid Inbound Parse, or any service that can POST a raw message. Each path is authenticated by your organization's inbound token, which you can rotate at any time, and delivers only to your own inboxes and verified domains. Shipped
Inbound routed by envelope recipient Mail is delivered to the inbox named in the SMTP envelope, not the To: header, whichever way it arrives, and redelivery is idempotent. Shipped
Raw message push Push raw MIME into any of your inboxes for imports and generic inbound webhooks. Shipped
Custom domains Add a domain and nofeb returns its DKIM, custom MAIL FROM and DMARC records, plus an inbound MX if you want nofeb to receive its mail, checks them, and emits domain.verified when they pass. Shipped
DNS setup made easy nofeb detects who hosts your domain's DNS and gives you the shortest path for that host. It can publish the records through the Cloudflare, DigitalOcean, Hetzner DNS, Vercel, Porkbun or Gandi API with a token used for that one call and never stored, and it never deletes a record. Every other host gets step-by-step instructions in its own terms, plus zone-file, Route 53 and Terraform exports. A live check shows each record's status, what was found and the likely mistake. Pointing your MX at nofeb is a separate choice, so publishing sending records never moves your mail. Shipped
Injection scoring and quarantine Every inbound message gets a prompt-injection score from 0 to 1 with its reasons: hidden text, invisible and look-alike characters, instructions aimed at an AI, role and chat-template tokens, tracking beacons, data-carrying links and encoded payloads. Above the threshold you set per organization, pod or inbox, the message is quarantined (or only flagged) for a person to release or delete, and its content stays out of the delivery event. Shipped
Sender trust on every message Each inbound message is labelled trusted, known, unknown or suspicious, with the reasons: its authentication results, your sender lists, look-alike domains, display-name spoofing, and whether the inbox has corresponded with the sender before. Shipped
Decisions on every message A default set of typed decisions on every inbound email (intent, urgency, needs-reply, automated, verification email, phishing, injection risk, requires-human and language) with a probability each, your own decision schemas beside them, human relabels that correct and retrain it, per-question accuracy and calibration stats, and confidence-gated sending for the send_when_confident tier. Decisions run without leaving nofeb by default; a third-party model sees mail only if you opt in. Shipped
Recipient and sender allowlists and blocklists Lists of addresses, domains or wildcards per organization, pod or inbox: who an agent may write to, and whose mail it reads. A blocked recipient is refused on every autonomy tier, a blocked sender's mail is rejected or quarantined, and you choose whether a send outside the allowlist is held for approval or refused. A check call predicts what a send or an inbound message would meet before it happens. Shipped
Rules and Sieve Declarative rules over sender, recipients, subject, body, headers, attachments, size, the security verdict, metadata and decisions, that label, move, flag, set metadata, quarantine, forward, auto-reply or reject. Test rules against a stored message with a per-rule trace, and import or export Sieve, with anything that cannot be mapped reported line by line. Shipped
HTTPS hooks in the pipeline Signed HTTPS callouts that change what happens: create or reject an inbox for an unknown recipient, accept, reject or quarantine a message before delivery, and allow, deny or hold a send, each with a timeout and a fail-open or fail-closed mode. Shipped
safe_body for untrusted mail The message's new text with zero-width and direction-altering characters, hidden HTML and forgeable quoted history removed, wrapped in a provenance header that tells the model whose words these are. Shipped
Forward to an agent Per-person forwarding addresses that bring someone's own mail into their agent's inbox byte for byte, accepting only mail traceable to the senders you allow. Gmail and Outlook forwarding-confirmation codes and links are recognised and surfaced, and each address can be rotated or revoked. Shipped