Skip to content
LookupOwl

Email health check

Check whether the mail for a domain is set up so that it arrives and cannot easily be faked, in one report with the fixes to make first.

Runs every check. Up to 15 seconds.

What the report covers

Mail depends on a handful of records that live in different places and are easy to get slightly wrong. Each has its own checker, and people rarely have the time to open all of them. This report runs the lot for one domain and puts the answer on one page.

  • Mail servers (MX): where mail for the domain is delivered, and which provider runs it when we can tell.
  • SPF: which servers may send as the domain, and whether the record stays under the limit of 10 lookups.
  • DKIM: the signing keys we can find among the common selectors, and whether any is shorter than 2048 bits or broken.
  • DMARC: what receivers are told to do with mail that fails, and whether the policy is enforced or only watching.
  • Blocklists: whether any mail server address of the domain is listed.
  • BIMI, MTA-STS and TLS-RPT: the optional records. Not having them is a recommendation, never a fault.

Each part gets one line and a status: pass, info, warning, fail, or "Couldn't check" when it did not answer in time. Every status has a word next to its icon, so the page reads the same without colour.

A level word and counts, not a grade

The top of the report says one of four things: healthy, good with things to fix, needs attention, or not working. The level is the lowest result among the parts, and the counts beside it say how many passed, how many are information only, and how many are warnings or failures. Nothing is weighted and there is no letter, because a letter invites you to chase the letter instead of the problem.

A part that is not set up, such as BIMI, counts as information and never lowers the level. A part we could not check never counts as a pass. If a fifth or more of the parts did not finish, the report says it is based on the checks that did.

Fix first

Under the answer you get up to three fixes in the order that matters most: a missing DMARC record before a short DKIM key, a broken SPF record before an unlisted logo. Each row says why it matters and what to do, and where a copy-ready record helps, it has a Copy button.

When we recognise the mail provider from the MX records, the advice is written for it. A 1024-bit key at Microsoft 365 is fixed in the Defender portal, not in your DNS, and the row says so. When we do not recognise the provider, the advice stays general rather than guessing.

Can someone send mail as this domain?

The last card answers that in plain words, from the SPF and DMARC records: no, partly, or yes. A domain with no DMARC record, or one set to p=none, can be faked and the fake will usually be delivered. A domain that publishes p=reject and holds subdomains to the same rule cannot. The email spoofing check page gives only this answer and runs faster.

Mailbox providers such as gmail.com are a special case. They publish p=none on purpose, because millions of people send through other services as @gmail.com, and a stricter policy would block real mail. The card says "Yes, and that is on purpose" rather than calling it a mistake.

What we touch and what we keep

We read public DNS records and the public policy files for MTA-STS and BIMI. We never open a connection to a mail server and we never send a message. The report is built for the one request, a result that is a few minutes old may be shown again to save the servers asking twice, and nothing is kept about you.

The whole report can take up to 15 seconds, because it runs several checks and some of them query many blocklists. Without JavaScript the page waits and shows the finished report. With it, you see the parts finish one by one.

Questions people ask

Do you send an email to test?

No. We never send a message. We read your DNS records and, for MTA-STS and BIMI, the public policy files those records point to.

Why is there no grade?

A grade invites comparison between very different setups and hides what to do. A level word, counts and three fixes say what to change, and a domain that sends no mail can still be healthy.

Why does it say BIMI, MTA-STS and TLS-RPT are not set up when my mail works?

They are optional extras. Mail arrives without them. We list them as information because they add something: a logo in some inboxes, a rule that senders must use encryption, and reports when delivery fails.

What does "Couldn't check" mean?

That part did not answer within the time limit, so we cannot say if it passes. It is never counted as a pass or a fail. Run the matching checker on its own, or try the whole report again in a minute.

Why are DANE and mail server TLS missing?

They need a test that connects to your mail servers on port 25. We have not built that yet, so rather than guess, the report says they are not included and leaves them out of the level.

Does it check subdomains?

Yes, you can enter one. We check that exact name, and when its DMARC policy comes from the parent domain, the report says so. Mail for a subdomain that has no records of its own often relies on the parent.

Can I use the result in a ticket or a chat?

The Copy summary row gives a plain text version with the level, the counts and the fixes. The Copy summary for an AI row gives one line with the fix names, which suits a prompt.

Use it from AI, the API or the command line

Every LookupOwl tool is also a free API and an MCP tool, with no key and no account. This one is email_health. Please be kind to the rate limits.

curl
curl "https://lookupowl.com/api/v1/tools/email-health?q=github.com"
MCP
claude mcp add --transport http lookupowl https://lookupowl.com/mcp