# Email tools for AI agents: 17 tools, $0.001 to $0.01 per call

Whizbrew has 17 Email tools. One call that returns data costs $0.001 to $0.01; a call that finds nothing or fails costs $0. They are part of one catalogue of 893 tools behind one MCP server and one API key ([all groups](https://whizbrew.com/tools?a=023d10ebbec1), [pricing](https://whizbrew.com/pricing?a=023d10ebbec1)).

## Example calls

`email.mx` ($0.001): The mail servers of a domain (MX records) by priority, the mailbox provider they belong to (Google Workspace, Microsoft 365, Zoho, Proton and others, from our own rule list), and whether the domain accepts mail at all (a null MX says it does not). A domain without mail servers is free.

```
curl -s -X POST https://whizbrew.com/v1/call/email.mx \
  -H "Authorization: Bearer $WHIZBREW_KEY" -H "content-type: application/json" \
  -d '{"domain":"github.com"}'
```

`email.spf` ($0.001): A domain's SPF record, parsed and checked: each mechanism, the policy for all other servers (-all, ~all, ?all, +all), every include followed with the number of DNS lookups it costs (the standard allows 10), void lookups, and the problems found (several records, too many lookups, +all, ptr, syntax). Up to 25 DNS questions a call. A domain without SPF is free.

```
curl -s -X POST https://whizbrew.com/v1/call/email.spf \
  -H "Authorization: Bearer $WHIZBREW_KEY" -H "content-type: application/json" \
  -d '{"domain":"github.com"}'
```

`email.dkim` ($0.001): A domain's DKIM keys. With a selector: that key, parsed (key type, RSA key size in bits, testing flag, withdrawn or not, the mail service it points to). Without one: about 60 common selector names are tried and the keys found are listed with the service each belongs to. DNS cannot list selectors, so a key under another name is not found this way; a domain with no key found is free.

```
curl -s -X POST https://whizbrew.com/v1/call/email.dkim \
  -H "Authorization: Bearer $WHIZBREW_KEY" -H "content-type: application/json" \
  -d '{"domain":"github.com"}'
```

Over MCP the same call is `call_tool` with `{"tool": "<name>", "arguments": {…}}`.

## Every tool in this group

| Tool | What it returns | Price per call that returns data | Per 1,000 calls |
|---|---|---|---|
| `email.mx` | The mail servers of a domain (MX records) by priority, the mailbox provider they belong to (Google Workspace, Microsoft 365, Zoho, Proton and others, from our own rule list), and whether the domain accepts mail at all (a null MX says it does not). A domain without mail servers is free. | $0.001 | $1 |
| `email.spf` | A domain's SPF record, parsed and checked: each mechanism, the policy for all other servers (-all, ~all, ?all, +all), every include followed with the number of DNS lookups it costs (the standard allows 10), void lookups, and the problems found (several records, too many lookups, +all, ptr, syntax). Up to 25 DNS questions a call. A domain without SPF is free. | $0.001 | $1 |
| `email.dkim` | A domain's DKIM keys. With a selector: that key, parsed (key type, RSA key size in bits, testing flag, withdrawn or not, the mail service it points to). Without one: about 60 common selector names are tried and the keys found are listed with the service each belongs to. DNS cannot list selectors, so a key under another name is not found this way; a domain with no key found is free. | $0.001 | $1 |
| `email.dmarc` | A domain's DMARC record, parsed and checked: policy (none, quarantine, reject), subdomain policy, percentage, report addresses, alignment modes, whether the policy is really enforced, and the problems found. A subdomain without its own record gets the one it inherits. A domain without DMARC is free. | $0.001 | $1 |
| `email.bimi` | A domain's BIMI record (the brand logo mail clients show next to its mail): logo address, certificate address, and whether the DMARC enforcement BIMI needs is in place. The logo is not downloaded. A domain without BIMI is free. | $0.001 | $1 |
| `email.mta_sts` | A domain's MTA-STS setup (it tells other mail servers to deliver only over an encrypted connection): the DNS record and the policy file with its mode (enforce, testing, none), mail server patterns and lifetime, and whether the patterns cover the domain's real mail servers. A domain without MTA-STS is free. | $0.001 | $1 |
| `email.tls_rpt` | A domain's TLS-RPT record: where it asks other mail servers to send reports about failed encrypted deliveries. A domain without one is free. | $0.001 | $1 |
| `email.senders` | Which email services a domain uses, from our own rule list over its mail servers, SPF includes (followed) and verification records: the mailbox provider and the sending tools (SendGrid, Mailgun, Mailchimp, Klaviyo, HubSpot, Amazon SES, Postmark, Zendesk, Salesforce and many more), each with the record that names it. Shows a company's email stack. A service that signs with DKIM only is not seen. | $0.002 | $2 |
| `email.domain_health` | One report on a domain's email setup: mail servers, SPF, DMARC, DKIM (common selectors), MTA-STS, TLS-RPT and BIMI, each pass, warn, fail or none with a one-line reason; a score from 0 to 100 with a grade by a weighting shown in the answer; and the issues ordered by severity, each with a plain fix. Blocklists are not part of it. | $0.003 | $3 |
| `email.syntax_check` | Checks the form of up to 50 email addresses without any network request: valid or not with the reason, the normalised form, local part and domain, and a likely typo of a big mailbox domain with the suggested fix (gmial.com to gmail.com). Syntax only: no mailbox is probed, a valid address can still not exist. | $0.001 | $1 |
| `email.address_type` | Sorts up to 50 email addresses without any network request: role address (info@, sales@, support@ ...), disposable mailbox domain, free mailbox provider (gmail.com, outlook.com ...) or business address, with plus tags, from our own lists. No mailbox is probed. | $0.001 | $1 |
| `email.verify` | Is this email address deliverable, as far as public records tell: a verdict (deliverable, undeliverable, risky, unknown) with a confidence and the reason, and each check on its own: the form of the address, role, throwaway and free-mailbox lists, a likely typo with the fix, whether the domain exists and publishes mail servers (and whose), whether a public profile is registered to exactly this address and, for a domain that signs in at Microsoft with its own passwords, whether a sign-in account with exactly this name exists (asked at the Microsoft sign-in page, no login; "found" is evidence, "not found" says little because aliases and groups have no account). For a domain that takes mail the answer also carries its SPF record and DMARC policy, and with "formats" the address formats the domain uses. The mailbox itself is NOT asked and catch-all domains are not detected: no mail server is contacted, so a well-formed address at a working domain is "unknown" unless public evidence confirms it. An "unknown" answer is free. | $0.001 | $1 |
| `email.verify_bulk` | email.verify for up to 25 addresses in one call, in the order given, with totals per verdict; each domain is looked up once. Same checks and the same limit: no mail server is contacted, so the mailbox itself is not asked. A list where every address comes back "unknown" is free. | $0.01 | $10 |
| `email.patterns` | How a company writes its email addresses ({first}.{last}@, {f}{last}@ ...), inferred from addresses that are public: the ones its own site prints, its security.txt, its press page and the authors of public commits in its GitHub account (the one its site links to, or the one given). Each format comes with how many addresses show it and a confidence, and every address used is listed with where it was found; role addresses (info@, press@) are listed apart. This infers from public evidence: it does not look anybody up in a contact database. A domain with no public address is free. | $0.003 | $3 |
| `email.guess` | The likely work email address of a named person at a company: up to 6 candidates for a first and last name, the formats seen in the company's public addresses first (see email.patterns), then the common ones. Each says what backs it: published (that person's address is public), confirmed by a public profile registered to it, likely (the company's main format), possible, or a plain guess; and each is run through email.verify. It infers and checks: it does not look anybody up in a private contact database, and no mail server is contacted, so the mailbox itself is not asked. An answer with nothing better than a guess is free. | $0.005 | $5 |
| `email.find` | The single best work email address of a named person at a company, with what backs it and an honest confidence (0 to 100). Two sources, asked at the same time and both shown. (1) A licensed contact index: the address it holds for the person's public professional profile. Give profile_url (a row of people.search has it), or leave it out and the profile is looked up by name and company; from a looked-up profile only an address at the company's domain counts. Such an address is best with status vendor_found and confidence 90: the index says it proves its addresses against the mail server and re-tests them every 30 days, we did not ask the mailbox ourselves. (2) Our own public evidence, in own_evidence and the alternatives: the addresses the company's domain publishes and the formats they show (its site, security.txt, press page, public commit authors: see email.patterns), web search results for the domain's addresses and the person's name (at most 3 searches), the domain's mail servers in DNS and email.verify's checks; its status is published, seen_in_public, format_backed or guess. NOT done by us: no login, and no mail server is asked (mailbox_checked is always false). Give the domain, or the company name (resolved with company.domain_from_name). A person who asked to be left out is never returned. Charged only when best is an address the index found or has a confidence of 40 or more from public evidence; a guess or no address is free. | $0.003 | $3 |
| `email.examples` | Real marketing emails by brand or theme (e.g. nike, welcome, abandoned cart, black friday) from the public gallery of Really Good Emails: each email's subject line, its page in the gallery and the address of a picture of the whole email, the closest matches first, with the gallery's count of matches. For seeing what a brand sends or collecting examples of one kind of email. The first page of results (up to about 100); the email's full HTML is on its page. One read of the gallery's search page, no login, no key, no database of ours. Nothing found is free. | $0.001 | $1 |

Inputs, output shape, health and measured success rate of each tool: [https://whizbrew.com/v1/tools](https://whizbrew.com/v1/tools) and [https://whizbrew.com/status#email](https://whizbrew.com/status#email).

## Start

- **An agent on its own**: read [https://whizbrew.com/agents.md?a=023d10ebbec1](https://whizbrew.com/agents.md?src=body-text&a=023d10ebbec1). It asks for a free quote (`GET https://whizbrew.com/v1/agent/guide?goal=<the task in plain words>`), then one request (`POST https://whizbrew.com/v1/agent/start?a=023d10ebbec1`) gives it a key that works at once with the free credit, with no email and no sign-up form. Its human claims the account later.
- **A chat app** (ChatGPT, Claude, Gemini): add the MCP server `https://whizbrew.com/mcp` as a custom connector and sign in. No key to copy.
- **An agent with a shell** (Claude Code, Codex, Cursor): read [https://whizbrew.com/llms.txt?a=023d10ebbec1](https://whizbrew.com/llms.txt?src=body-text&a=023d10ebbec1). The agent starts a pairing with one request, its user approves with one click, and the agent holds its own key.
- **Plain HTTP**: create an account at [https://whizbrew.com/app#signup](https://whizbrew.com/app#signup), make a key, then `POST https://whizbrew.com/v1/call/<tool>` with `Authorization: Bearer <key>`.
