Make CLI: A new scriptable way to manage automations, built for developers and AI agents
If you, your dev team, or your AI coding agents live in scripting, CI/CD, and Git, this one is for you! 
Today, weβre introducing Make CLI: an open-source command-line interface that gives developers and AI agents full programmatic access to Make directly from the terminal.
What you can do with Make CLI
Manage your entire Make setup from the terminal: scenarios, connections, data stores, hooks, executions, and more
Give AI agents full Make access: AI coding agents like Claude Code or Cursor can manage automations via terminal with no token overhead or context pollution
Run automations natively in your CI/CD pipeline: integrate Make directly into your automated pipelines and scripting workflows
How Make CLI works
Make CLI is a tool you install on your computer or server. Once installed, you authenticate with your Make account and run commands directly from the terminal β no browser, no UI. Each command maps to an action in Make: listing your scenarios, activating one, checking executions, and so on. Itβs Make, but through text commands instead of clicks.
When to use Make CLI
Make now offers multiple ways to build and manage your Make automations. Mix and match based on how you work:
-
πππ€π π©π₯ππππ¨π«π¦ for visual, no-code building and orchestration. -
πππ€π πππ for AI assistants with structured tool calling -
πππ€π πππ ππ¨π¨π₯ππ¨π±ππ¬ for AI assistants when curated tool selection matters -
πππ€π πππ for programmatic control inside your own app -
πππ° πππ€π πππ for AI coding agents, scripting, and CI/CD
Same platform. Same power. Different ways to automate for every person and AI agent on your team.
Make CLI or MCP: Which to use for AI?
Use MCP when your AI is an assistant. CLI when your AI agent is a developer.
Both Make MCP and Make CLI give access to Make, but they are built for different kinds of AI.
Make MCP is native to AI chat assistants: conversational tools like Claude or ChatGPT, where a human is in the loop. MCP provides structured, native tool calling that fits how assistants work.
Make CLI is native to AI coding agents: tools like Claude Code, Cursor, or Codex that already live in the terminal and execute commands autonomously.
For AI coding agents, CLI is the natural fit because:
-
It fits into the workflows AI coding agents already use with version control, code review, and CI pipelines
-
There are no tool schemas loading into the agentβs context window; instead, it just runs commands, which reduces token spend and minimizes agent context overhead
-
Output is predictable and reviewable, like any other code change
The simple rule: Use MCP when your AI is an assistant. CLI when your AI agent is a developer.
Get started
Make CLI is available to all users as an open-source tool.
To install the Make CLI and run your first commands, follow the Make CLI documentation
Use with caution. Make CLI is intended for developers and advanced users. Commands can create, modify, and delete your scenarios, connections, data stores, and other resources. Always test in a non-production environment before considering using it in live workflows.
Technical highlights
-
Flexible installation: Install via Homebrew, npm, a precompiled binary, or a Linux package to fit your development environment
-
CI/CD and scripting support: Integrate Make into automated pipelines and non-interactive workflows using environment variable authentication or command flags
-
Multiple output formats: Choose between JSON, compact, or table formats to match your scripting needs and integrate with existing toolchains.
Helpful resources
Make CLI:
Make MCP:
Make MCP Toolboxes:
Make SDK:
Let us know your thoughts!
And as always, happy automating 
Valery from Make