We're excited to share something experimental from Postmark Labs: our new Model Context Protocol (MCP) server for Postmark. Think of it as giving AI assistants like Claude direct access to your email infrastructureâno more copying and pasting API calls or explaining how to send emails. Your AI can now actually send them.
Here's what makes this cool: instead of asking your AI assistant "How do I send a password reset email?" and getting instructions, you can now say "Send a password reset email to john@example.com" and it actually happens through your Postmark account.
The MCP server exposes four core Postmark capabilities as tools that AI assistants can use directly:
Our MCP server currently works with a single Postmark server (we had to start somewhere!). Once set up, you can have natural conversations with AI assistants about your email:
"Send a test email to the team about tomorrow's deployment"
"Show me our delivery stats for welcome emails this month"
"List all our email templates"
The AI handles all the API complexity while you focus on what you want to accomplish.
This is a Labs release, which means it's designed for tinkering and testing. We're sharing it early because we want your input on where to take it next. Some ideas we're exploring:
But we'd love to hear your ideas too. What email tasks would you want to delegate to an AI assistant?
Ready to give your AI assistant email superpowers? The Postmark MCP server is available now:
Try the MCP server on GitHub â
Read the full announcement and setup guide â
The setup is straightforwardâjust clone the repo, configure your Postmark credentials, and connect it to Claude Desktop, Cursor, or any MCP-compatible AI tool.
This is just the beginning of bringing email and AI together in meaningful ways. We'd love to hear about your experiments and creative use cases. Share your feedback and help us figure out what comes next!
