Acceptable Use Policy

1. The principle

MCPmailer is for conversational email: an agent that replies to people who wrote to it, follows up on a thread it started with a real person, and stops when asked. Everything below follows from that.

It is not a bulk-mail platform. The product is built so that replying is the cheapest and easiest thing an agent can do, and mass sending is the hardest. Trying to make it into a bulk sender will fail on quotas long before it becomes a policy question.

This policy is part of the Terms of Service and applies to everything sent through the service, including everything your agents do on their own.

2. What you may not send

• – Unsolicited bulk email, in any volume, to people who did not ask to hear from you and have no relationship with you.
• – Mail to purchased, scraped, rented, or otherwise harvested address lists.
• – Mail to anyone who has unsubscribed, bounced permanently, or complained.
• – Phishing, credential harvesting, malware, or anything designed to deceive a recipient about who is writing.
• – Impersonation of a person, company, or authority, including sending from a domain you do not control.
• – Illegal content, or content that is defamatory, harassing, threatening, or sexually exploitative.
• – Content that infringes copyright, trademark, or other rights.
• – Regulated or high-abuse categories that consistently damage sending reputation: cryptocurrency and token promotion, gambling, adult content, payday and high-interest lending, pharmaceutical sales, and get-rich-quick or multi-level marketing offers.

3. What you may not do to the service

• – Evade quotas, rate limits, suppression lists, or abuse controls, including by spreading the same campaign across multiple workspaces, mailboxes, or domains.
• – Create accounts automatically in bulk, or create a new account to escape enforcement on an old one.
• – Vary message bodies mechanically to defeat duplicate detection.
• – Probe, scan, or attack the service, or try to access another customer's data.
• – Use the tunnel service to expose anything that would breach this policy if it were mail, or to proxy traffic unrelated to your agents.
• – Resell the service without a written agreement with us.

4. Quotas

Quotas depend on your plan and on your sending history, and they rise as that history stays clean. They are enforced per mailbox and per workspace, in real time.

On the free tier: up to 3 recipients per message, 10 cold sends per day, 10 new recipients per day, and 200 replies per day. On a new paid workspace: 5 recipients per message, 50 cold sends and 50 new recipients per day, and 1,000 replies per day. Once a workspace has an established clean record: 5 recipients per message, 500 cold sends and 500 new recipients per day, and 10,000 replies per day.

Cold sending, meaning mail to someone who has not written to you first, requires a payment method on file. This is deliberate friction: it costs nothing to a legitimate business and a great deal to someone spinning up throwaway accounts.

A reply to an inbound message is never counted as cold, which is why the reply limits are an order of magnitude higher.

5. How enforcement works

Enforcement is automatic and based on behaviour, not on the content of your mail. We do not read messages to decide whether to act.

A workspace is throttled when its rolling 24-hour bounce rate passes 2%. Throttling lifts by itself once 24 hours pass with no further bounces or complaints, so a bad import that you fix costs you a day, not an account.

A workspace is paused when its bounce rate passes 4% or its complaint rate passes 0.05%, which is half the level that puts a sending platform at risk with mailbox providers. A pause does not lift by itself; you have to contact us and explain what changed.

Two other tripwires fire on patterns rather than rates: a sudden burst of cold sending more than five times your recent average within an hour, and the same message body reaching more than 20 distinct new recipients within 24 hours. Both indicate a campaign, which is not what this service is for.

We may also suspend an account immediately, without notice, where recipients are at real risk, or where our sending reputation is threatened. We will always tell you what happened and, where the cause is fixable, what would restore the account.

6. What we expect of you

• – Send only to people who have a reason to hear from you, and be able to show what that reason is.
• – Honour unsubscribes immediately. We add them to a suppression list automatically, but do not go looking for ways around it.
• – Identify yourself honestly. Say who is writing and, where the law requires it, that a message comes from an automated agent.
• – Verify the domains you send from and keep their DNS records in place.
• – Keep your lists clean. Repeated hard bounces are the fastest route to a throttled workspace.
• – Supervise your agents. An agent that starts emailing strangers is your responsibility to notice and stop.

7. Reporting abuse

Report abusive mail sent through MCPmailer using the contact form, choosing "Abuse report". Include the full message with headers if you can. We investigate every report, and we act on the ones that are founded.

If you are a recipient trying to stop mail, the unsubscribe link in the message is faster than we are: it takes effect immediately and permanently for that sender.

8. Changes

We update this policy as abuse patterns change. Material changes are announced by email at least 30 days in advance, except where a change is needed immediately to stop active abuse.