Want a free personalized recommendation for your software stack?Get yours
That Marketing Buddy
Home/MCP/Postmark
Postmark

Postmark MCP Server

Yes. Postmark annotates every tool with readOnlyHint and destructiveHint, so a client can auto-approve a bounce lookup while still stopping to ask before a batch send.

OFFICIAL MCPPublic APIAuth: Postmark server token, supplied as an environment variable to the local processLLMs: Claude (Desktop/Web)
MCP status verified 2026-08-06Agent-readiness 7/10
Data from Buddy's database, auto-rechecked weekly

What is the Postmark MCP server?

Postmark ships an official MCP server, published by ActiveCampaign as @activecampaign/postmark-mcp. It runs locally rather than as a hosted endpoint: you install it via npm and point your client at the local process with a Postmark server token in the environment. There is no URL to connect to, which is the correct architecture for something holding a sending credential.

It exposes 24 tools across email sending (single and batch), templates (full CRUD plus validation), message search, delivery diagnostics, bounces, suppressions, stats, server info and webhooks. That covers both halves of transactional email: pushing messages out and working out why one did not arrive.

The standout is the annotation work. Every tool carries readOnlyHint and destructiveHint markers, so a supporting client can auto-approve safe reads and force a confirmation before anything mutating or destructive. Postmark also masks email addresses in its logs by default, enforces HTTPS, and offers a domain allowlist for webhook registration.

For the rest of the picture see the Postmark review or full pricing breakdown.

Inside Claude / Cursor / Codex

What you can build with the Postmark MCP

Illustrative workflow examples. Exact tool names, parameters, and outputs come from Postmark's MCP server schema, so check the docs for the live surface.

postmark-mcplive
>What's the open rate on my last 5 campaigns?
postmark.list_campaigns({ limit: 5 })
LAST 5 CAMPAIGNS · OPEN RATE
Black Friday Teaser42.1%
Product launch38.5%
Weekly digest #1227.0%
Welcome flow step 261.3%
Re-engagement attempt9.8%
Suggest A/B subject lines for the bottom two?
postmark-mcplive
>Build a 3-email welcome series for new ebook subscribers
postmark.create_automation({ trigger: "list_added" })
AUTOMATION CREATED · 3 EMAILS
Day 0ebook delivery + intro
Day 2case study + soft CTA
Day 5product demo + 14-day trial
Review drafts before activation?
postmark-mcplive
>Which segments are growing fastest this quarter?
postmark.segment_growth({ period: "Q4" })
TOP 3 GROWING SEGMENTS
Trial → Paid converters+187%
EU subscribers+62%
Re-engaged churned+34%
TMB SCORE FOR PostmarkMCP: OfficialAgent fit: 7/10
Read review →

Access, auth, and limits

Auth is a Postmark server token passed as an environment variable to the local process, so the credential stays on your machine and never transits a third party. Node.js v20 or higher is required. Because it is stdio rather than a remote endpoint, there is no MCP URL: any directory listing showing one for Postmark is describing the GitHub repository, not a server you can reach.

The operational details are unusually well handled. Logging is structured JSON to stderr with optional file persistence, and email addresses are partially masked by default rather than as an opt-in. Webhook registration enforces HTTPS and supports a domain allowlist, which closes the obvious hole where an agent could be talked into pointing your webhooks somewhere else.

Where it works and what to build

Claude Desktop and any stdio-capable MCP client work. Clients that understand tool annotations get the most out of it, because that is what turns readOnlyHint into an auto-approved read and destructiveHint into a confirmation prompt.

The use that pays for itself is delivery forensics. When a customer says they never got the receipt, an agent can search messages, pull the delivery diagnostics, check bounces and inspect suppressions in one pass, and tell you whether the address hard-bounced weeks ago. That is a multi-tab job in the dashboard and a single question here. Sending is available too, but the read path is where the daily value sits.

Which AI clients can use Postmark MCP?

Not every AI assistant supports MCP natively. Per-client picture for Postmark:

ClientSupportNotes
Claude (Desktop/Web)Native MCP

Does Postmark MCP cost extra?

Included with subscription

Postmark ships an official MCP server that uses your existing subscription credentials.

  • No separate MCP licence to buy. You pay for your Postmark plan, the MCP server queries that account.
  • Whether it's gated to a specific tier wasn't extracted automatically — verify the access requirements on the MCP page or pricing docs.

Postmark also publishes a REST API

Postmark publishes an API, but per-tier gating wasn't extracted automatically.

Auth method
SDK languages
OpenAPI spec
Webhooks
Structured outputs
See Postmark API page

What you can do with Postmark from an AI agent

Three shapes of agent workflow. Pick the row that matches how you actually use Claude / Cursor / Codex day to day.

INDIE / SOLO

Skip the dashboard, ask Claude

  • "What campaigns underperformed this week?" answered in seconds
  • Quick one-off questions without leaving the editor
  • Pipe Postmark data into your personal notes / docs without writing a script
Effort: 30 seconds — paste the config, restart Claude
AGENCY

Run client analysis at agent speed

  • Hand the MCP server to your team — every consultant can ask Claude about any client's Postmark account
  • Multi-account workflows without context-switching the Postmark UI
  • Reports you'd normally spend hours on, done before standup
Effort: An afternoon — set up auth per client
IN-HOUSE TEAM

Production agent workflows

  • Embed in your internal Codex / Cursor agents that triage Postmark data daily
  • Combine with other MCP servers in a multi-step agent chain
  • Constrain agents to read-only roles to keep production Postmark safe
Effort: Days — security review + role config

Limits and gotchas

  • Official MCP server — published by Postmark themselves, so it should track product changes.
  • Fallback: the REST API works for any flow the MCP server doesn't cover.
  • Rate limits: MCP calls count against your underlying Postmark API quota. Burst usage from a curious agent can drain a daily allowance fast.
7/10

Agent-readiness verdict

The most safety-conscious local MCP server in the email category. Tool annotations plus log masking plus a webhook allowlist add up to a server built by people who thought about what an agent could get wrong.

Moderate agent-fit with official MCP support. Postmark offers an official MCP server and documented public API, giving agents direct access to transactional email workflows. The lack of OpenAPI specs, SDKs, and webhook support limits deeper integrations and real-time event handling.

Scored by Joonas (TMB).

Postmark MCP & API FAQ

Is the Postmark MCP server official?+
Yes — an official, vendor-maintained server at null.
Can I use the Postmark MCP with Cursor or Codex?+
Yes — MCP is a protocol, not Claude-specific. Any MCP-capable client (Cursor, Codex, Windsurf, Zed) can use the same server config with minor path adjustments.
Does Postmark have a public API?+
Yes — public REST API at postmarkapp.com/developer.

Sources

Data verified by Joonas on the dates shown. MCP server status auto-rechecked weekly.

Joonas Rotko
Joonas RotkoAuthor & Founder of That Marketing Buddy

10+ years in digital marketing. I review marketing software for AI-stack fit: real pricing, MCP and API support, and how cleanly each tool drops into an AI agent workflow, cross-checked against verified data and real user feedback.

Postmark

Try Postmark
WEEKLY NEWSLETTER

The marketing tools worth your money, scored for AI agent readiness

Every week I score the tools for MCP, API, and agent-readiness, then email you the ones that actually deliver, what plugs into Claude and Cursor, the pricing changes, and what to skip. No fluff, no daily spam.

Join for free. One email a week, unsubscribe anytime.