Activepieces MCP
Activepieces is an open-source, AI-first automation platform with 280+ customizable TypeScript integrations for secure, no-code workflow building.
How to Install and Use Activepieces MCP
If you want to automate tasks easily and connect many apps, Activepieces MCP is a great tool. It works with powerful AI and is simple for developers and non-coders alike. Below are clear steps to get Activepieces MCP running and to use it with example code.
Step 1: Set Up Your Environment
Before installing Activepieces MCP, you need to prepare your computer. It runs best with Node.js, so first make sure you have Node.js installed. You can download Node.js from nodejs.org.
Once Node.js is ready, you'll want to get the Activepieces MCP source code or install it through npm.
Step 2: Install Activepieces MCP
Here's how to install Activepieces MCP. Open your terminal or command prompt and run this command to install it globally:
npm install -g @activepieces/cli
This command gives you access to the Activepieces command-line tools needed to manage MCP servers.
Step 3: Initialize a New Project
After installation, create a new Activepieces MCP project. Start by making a folder for your project, then move there in the terminal and run:
activepieces init
This command sets up a fresh project with the necessary files. It creates a simple starting point for building your MCP integrations.
Step 4: Develop Your Piece
Pieces are what connect to applications inside Activepieces MCP. You write them in TypeScript. Here’s an example of a simple piece action using TypeScript:
import { createAction, Property } from '@activepieces/pieces-framework';
export const exampleAction = createAction({
name: 'example_action',
description: 'This is a simple example action',
props: {
message: Property.ShortText({
displayName: 'Message',
required: true,
}),
},
async run(context) {
const message = context.propsValue.message;
console.log('You sent:', message);
return message;
},
});
This code defines an action that takes a message and prints it.
Step 5: Run Your MCP Server Locally
To test your pieces or create your own integrations, run the MCP server locally. Use this command inside your project folder:
activepieces dev
This will start the Activepieces MCP server with hot reloading. When you make changes to your piece code, it reloads automatically so you can see results quickly.
Step 6: Connect and Use Pieces
Once your MCP server is running, you can use it with LLM tools like Claude Desktop or Cursor. They communicate with your pieces to run automations powered by AI.
You can also share your pieces by publishing them as npm packages so other users can install and use them.
By following these steps, installing and using Activepieces MCP becomes easy and practical. The tool is built to let you customize workflows quickly using TypeScript, while giving you AI-powered automation abilities. Start with the installation commands, create your pieces with simple TypeScript code, and run your local MCP server to explore building smart automations!