Integration
MCP for Postmark: the straight answer, and the one install rule that matters.
You run transactional email on Postmark because it's fast and it lands. But the newsletter you bolted onto a broadcast stream has started drawing spam complaints, the bounce rate crept up after the last product launch, and you half-suspect someone's been sending bulk mail out of the transactional stream where it poisons your best reputation. So you open five tabs, Message Streams, Bounces, Suppressions, Sender Signatures, the Stats page, and start reconciling by eye. Ask Claude for help and it gives you good advice about all of it. Generic advice. It hasn't seen your streams or your bounce feed. If you searched "MCP for Postmark" hoping to close that gap, this page owes you the honest version of what that search finds, including the one place a careless install genuinely bit people.
What "MCP" actually means here.
MCP stands for Model Context Protocol, an open standard from Anthropic that lets Claude reach outside the chat window. It's a plugin system where the plugins give Claude real capabilities. Two facts before any pitch. First: an official Postmark MCP exists. ActiveCampaign, Postmark's owner, publishes it at github.com/activecampaign/postmark-mcp, 24 tools running locally with your Server token. The one rule: install it from that repository and nowhere else, because in September 2025 a fake npm package using the same name backdoored roughly 300 organisations by silently BCCing their email. Second: Orbit ships no Postmark API tools. Nothing in Orbit takes a Postmark token. What Orbit is, is the deliverability and lifecycle layer around the sending: strategy, build, QA, scoring, all running without a credential from anyone. The full walk-through of what an MCP is and how it changes the work lives on the MCP for marketing page →.
What works beside Postmark today.
No hand-waving. Each row names who provides it: the live connection is the official postmark-mcp from ActiveCampaign's GitHub, the deliverability layer is Orbit's and runs locally with no credential, the export rows say exactly how your data arrives, and the row that doesn't exist says so.
Connect Postmark to Claude live
Works todayThrough the official postmark-mcp from ActiveCampaign, Postmark's owner: 24 tools, running locally on your machine over stdio with your POSTMARK_SERVER_TOKEN. Install it only from github.com/activecampaign/postmark-mcp: a fake npm package using the same name backdoored roughly 300 organisations in September 2025.
Plan suppression and bounce strategy
Works todayOrbit's side, no token needed. Which suppressions to mine, how to separate list-hygiene problems from authentication problems, where the complaint-rate thresholds sit now that Gmail and Yahoo enforce them. The protocol a deliverability owner uses weekly.
Reason about SPF, DKIM and DMARC
Works todayPaste your DNS records and Claude reviews the whole picture: whether SPF and DKIM align with your From domain, and where your DMARC policy sits on the path from monitoring to enforcement. The records are yours to fetch; the reasoning is the hard part.
Keep streams honest: transactional vs broadcast
Works todayPostmark separates transactional and broadcast streams precisely so bulk mail can't drag down the reputation that gets your receipts delivered. Orbit carries the stream-hygiene protocol: describe your setup, or read it through the official server, and Claude flags the classic mistake before it costs inbox placement.
Build and QA transactional templates
Works todayOrbit's local pipeline: copy composed, MJML compiled to clean, dark-mode-safe HTML with your template placeholders left intact, and the render gate run before anything ships. You paste the result into Postmark's UI: one manual step, named plainly.
Score subject lines and preheaders
Works todayHeuristic scoring built from what actually moves opens and what trips spam filters, run before the send, not explained after it.
Analyse your bounce and stats exports
Partial: named limitNo live read from Orbit. It has no connection to your Postmark account. Export the bounce list or a stats report and Claude reasons over the real numbers. Genuine analysis; the transport is your clipboard, or the official server if you've connected it.
Call Postmark's API from Orbit's own tools
Not shippedNot shipped. No Orbit tool reads or writes a Postmark account today, and with the official server already covering the live work, that gap costs you nothing. If Orbit ships its own Postmark tools, this page changes the same day.
The short version: the live connection exists and it's official, from Postmark's owner, on GitHub. The deliverability judgement exists and it's Orbit's: local, key-free, ready this afternoon. Orbit's own pipe into the Postmark API doesn't exist, and with the official one already there, it doesn't need to.
The transactional audit it was built for.
This is where the layer earns its place. Ask Claude to run the checks a deliverability owner does by hand every quarter, working from your DNS records, your Postmark exports, and the official server's reads if you've connected it.
- Stream hygiene: transactional vs broadcast. Postmark keeps transactional and broadcast mail on separate
Message Streamsprecisely so a bulk send can't drag down the reputation that gets your receipts and password resets to the inbox. Orbit carries the protocol for keeping that split honest and flags the classic mistake, newsletter mail on the transactional stream, before it costs you inbox placement on the mail that actually matters. - Bounce and suppression review. Export your bounces and broadcast-stream suppressions and Claude sizes them against your send volume. A rising hard-bounce or complaint rate surfaces as a number, not a hunch, and list-hygiene problems get separated from authentication problems, which look the same from the inbox but need different fixes.
- Sender authentication: SPF, DKIM, DMARC. Paste the records from your DNS console and Claude reviews the whole picture: whether your DKIM and Return-Path align with your From domain, and where your
_dmarcpolicy sits on the path fromp=nonethroughquarantinetoreject.
None of this needs Orbit to hold a token. The records are public, the exports are two clicks in Postmark's UI, and the reasoning is the part that was never going to come from a key anyway.
How to set it up.
Orbit first, official server second if you want the live reads.
- Install Orbit: no Postmark step exists. It's free, and there is no token field because no Orbit tool would use one. The deliverability protocols, the pipeline, the QA gate, and the scoring run immediately. Your Postmark tokens stay in Postmark.
- Add the official postmark-mcp for live access. Clone or install it from github.com/activecampaign/postmark-mcp, ActiveCampaign's own repository, the only source worth trusting with a Server token, and configure it with your
POSTMARK_SERVER_TOKEN. It runs locally over stdio, so the token stays on your machine.
If Orbit ships direct Postmark tools later, the connection step will appear here, and it will start read-only, because that's the posture everywhere else Orbit connects.
What's deepest on Braze, and why we say so.
Orbit's craft applies to every major ESP, but its deepest single-platform integration is Braze. A handful of Orbit's tools are built around concepts that only exist in Braze, Canvas flows, Content Blocks, its media library, and those stay Braze-named on purpose. Faking a Postmark equivalent would just be a worse version of a thing Postmark's model doesn't have.
For Postmark, that trade barely bites: the portable work, the deliverability protocol, the stream hygiene, the template HTML, the QA gate, the numbers, is exactly what a Postmark-based developer or deliverability owner needs. And if you later graduate to a lifecycle-native platform, the tooling and the discipline travel with you.
The part that was never about an API.
Everything Orbit brings here is knowledge and local tooling, which is exactly why it works without a connection. Orbit ships 88 lifecycle skills, structured protocols for the actual jobs: auditing deliverability, composing an email, laddering a win-back, scoring a subject line. Those are ESP-agnostic; they work the same whether the account underneath is Postmark or anything else.
Underneath sit 136 tools, an MJML pipeline that compiles to clean, dark-mode-safe HTML, a render-and-QA gate that catches the clipping and inbox breakage before you send, and the calculators a real test needs: sample size, holdout, replenishment timing. Postmark gives you the send and the deliverability rails. Orbit is the practitioner sitting next to them, reasoning about the same numbers you do.
Is this for you?
This is for the developer or deliverability owner who runs Postmark and wants a sharper practitioner beside it, the person watching the bounce feed, keeping the streams clean, and making sure the sender signatures stay confirmed. If that's your week, the working setup is Orbit for the judgement and, when you want live reads, the official postmark-mcp from ActiveCampaign's GitHub for the hands. It's a weaker fit if you expected the live pipe to be Orbit's. It isn't, and we'd rather point you at the official one than build an unofficial copy of it. What you get is a Claude that works like a deliverability practitioner instead of a search engine, and a straight answer about what's real and where to install it from, which, given what happened to this server's name on npm, is worth more here than usual.
Try it.
Every Orbit plan includes every skill and tool, on every ESP. Install it and the whole deliverability layer runs the same afternoon: no Postmark token, nothing to revoke later. Add the official server from ActiveCampaign's GitHub when you want the live reads.
Run a different ESP?
The same honest treatment, per platform:
Frequently asked.
What is an MCP for Postmark?
MCP stands for Model Context Protocol, an open standard from Anthropic. It gives Claude real tools instead of general knowledge. For Postmark, two honest facts. First: an official MCP server exists. ActiveCampaign, Postmark's owner, publishes it at github.com/activecampaign/postmark-mcp, with 24 tools that run locally against your Server token. Second: Orbit ships no Postmark API tools; nothing in Orbit takes a Postmark token. What Orbit is, is the deliverability and lifecycle layer around the sending: suppression and bounce strategy, SPF, DKIM, and DMARC reasoning, a local email pipeline with a pre-send QA gate, and every part of it runs without a credential from anyone. Orbit is not an official Postmark product; that title belongs to ActiveCampaign's server, and Orbit is the independent MCP that sits beside it.
Is there an official Postmark MCP?
Yes. ActiveCampaign publishes it at github.com/activecampaign/postmark-mcp. It carries 24 tools, runs locally on your machine over stdio, and authenticates with your POSTMARK_SERVER_TOKEN. One rule matters more than the feature list: install it only from ActiveCampaign's own GitHub repository, never from an unverified npm package name. In September 2025 a fake package called postmark-mcp, unrelated to ActiveCampaign, backdoored roughly 300 organisations by silently BCCing their email. The official server is good; the name has been counterfeited once already.
What does Orbit actually do for a Postmark sender?
The work around the send, which is most of the deliverability job. Orbit carries the practitioner's protocol: suppression and bounce strategy, complaint-rate thresholds now that Gmail and Yahoo enforce them, stream hygiene, keeping bulk mail off the transactional stream that carries your receipts, and SPF, DKIM, and DMARC reasoning from the DNS records you paste. Underneath sits a local email pipeline: MJML compiled to clean, dark-mode-safe HTML with your template placeholders left intact, a render gate that catches clipping and dark-mode breakage, and subject-line scoring. None of it asks for a token.
Can Orbit read my bounces or stats?
Not itself. Orbit has no connection to your Postmark account. Two paths work today. Live: connect the official postmark-mcp from ActiveCampaign's GitHub and ask Claude to use its tools alongside Orbit's reasoning. By export: pull your bounce list or a stats export from Postmark's UI, hand it to Claude, and the analysis runs on your real numbers. It separates list-hygiene problems from authentication problems, which look identical from the inbox but need different fixes.
Can Orbit build my Postmark templates?
It builds the HTML; you do the paste. Orbit's pipeline composes the copy, compiles MJML to clean, dark-mode-safe HTML, keeps your template placeholders intact, and runs the QA gate. Then you drop the result into a template in Postmark's UI yourself. That's one manual step more than an API push, and we'd rather name it than imply a write path Orbit doesn't have. The build and the QA, the parts that actually take the time, are done for you.
Is Postmark the right home for a serious lifecycle program?
Postmark is strongest as a transactional-email engine: the layer that gets receipts, password resets, and notifications to the inbox fast, with a reputation kept clean by separating transactional and broadcast streams. Its broadcast streams handle bulk and newsletter sending with proper suppression and unsubscribe management, which covers a lot of SMB lifecycle work. Where it strains is deep, branching, behaviour-driven journeys and true incrementality testing across a large audience; if that's the week you're living in, a lifecycle-native platform like Braze is the graduation. Orbit's discipline applies to both, deepest on Braze, so the tooling and the craft travel with you if you move.