Back to overviewWe recently released the Lettermint MCP server, which lets you ask your AI assistant about your email delivery. Prefer to do things yourself in a terminal? Then the Lettermint CLI is for you.
Sending a transactional email is now a single command, no dashboard required. Want to know what happened to a message? Look it up and see its events. And you can test your webhooks right on your own machine. The CLI runs on macOS, Linux and Windows.
What is the Lettermint CLI
CLI stands for command line interface, a program you control from your terminal. You type a command and get the result back as text or JSON. Many developers prefer this over a dashboard because they can stay in their IDE instead of switching to a browser.
A command is also something you can reuse. Put it in a script or run it in CI. An AI agent can use it too. That is why tools like Git and Docker have had a CLI for years, and now Lettermint has one as well.
What you can do with the Lettermint CLI
Anything you would normally open the dashboard for, you now type as a command. Sending an email, for example, or checking why a message has not been delivered yet. You can also send a test version of your newsletter to your own inbox before it goes out to your whole list. The answer shows up in your terminal and you carry on with your work.
Here is what the CLI can do:
- Send emails and view the events of a message
- Export the content of a sent message as HTML, text or raw
- Create projects and manage routes
- Add domains, verify them and assign them to a project
- Manage webhooks and receive them locally on your own machine
- Switch between teams using profiles
- Export an agent skill for your AI agent
Test webhooks on your own machine
Testing webhooks locally used to be a hassle. You want to see whether your application handles a bounce correctly, or what happens when a reply comes in on an inbound email. But your computer is not reachable from the internet, so you set up a tunnel with a tool like ngrok. That worked, but it was one more tool to install and keep an eye on.
With the Lettermint CLI you start a listener instead. Lettermint then forwards the events straight to your application on localhost. No tunnel, and your existing webhooks keep running as usual.
lettermint webhooks listen --project PROJECT_ID \
--forward-to http://localhost:3000/webhooks/lettermint
Every bounce, delivery or inbound email shows up in your terminal, along with the status your application returned. If something in your code is off, fix it and replay the same event. No need to send a new email.
Scripts and AI agents
Anything you type in your terminal can go in a script. Check in CI whether a domain is still set up correctly, or create a project for a new client while you are in the terminal anyway.
Working with an AI agent like Claude Code or Codex? It can use the CLI too. The CLI ships with an agent skill that explains how the commands work and what an agent should watch out for. Export it to a folder in your project:
lettermint skills export --output ./lettermint-skills
Point your agent at that folder and it knows how to send an email or check what happened to one.
Secure login
You do not need an API token. You connect the CLI to your account as an application, just like the MCP server. Log in through your browser and approve access, and from then on your team permissions apply to every command.
lettermint auth login
Under Applications in your account you can see which connections exist and revoke them when you stop using a computer. Working with several teams? Log in once per team and switch with lettermint profiles use.
Available now
The Lettermint CLI is available for macOS, Linux and Windows. Install it with Homebrew on macOS, with the install script on Linux, or through PowerShell on Windows. Your email still goes through our European infrastructure, the CLI does not change that. You can find all commands, options and examples on GitHub (opens in a new window).