In my latest exploration of Make.com, I focused on its exciting new feature, the Model Context Protocol (MCP). This enables users to connect external agents, such as Claude Desktop, directly to their Make.com accounts. The beauty of this feature lies in its ability to let these agents discover, understand, and execute scenarios without requiring complex setups. I’ll walk you through the process of getting started, how it works, and share practical examples along the way.
MCP in Make.com Explained
The Model Context Protocol, or MCP, acts like a universal connector for tools and services. With this protocol, AI agents can interact with various tools without the need for individual integrations. It’s designed to make automation straightforward and efficient. I found it remarkably easy to set up Claude Desktop to connect with my Make.com account. In just a few minutes, I could enable Claude to access various scenarios, such as creating tasks in ClickUp, generating AI images, and posting to WordPress.
Once the connection is established, Claude can instantly identify any scenarios marked as “on demand.” This means you won’t have to manually wire up each integration. For example, I asked Claude to add a new task to my ClickUp list. It seamlessly called the ClickUp scenario and responded with a URL. When I opened that link, I saw that the task was successfully created. Then, I prompted it to generate an image, and it responded with a file from a separate scenario dedicated to image generation.
Setting Up Make.com MCP
To get started with the MCP feature, you’ll need to set up your Make.com account for API access. This involves a few straightforward steps:
- Go to your profile and navigate to API MCP access.
- Add a new token by selecting the MCP token option and pressing add.
- Copy the URL provided by Make.com; this URL allows Claude to communicate with your account.
When Claude connects to Make.com, it can discover all the scenarios you’ve set up. This is where the real advantage of the MCP protocol comes into play. As you add new scenarios to your account and mark them as “on demand,” Claude can find them without any additional configuration. This flexibility extends beyond Claude Desktop. Other external tools with MCP clients can also be integrated into Make.com, provided they understand the protocol.
Understanding Scenario Inputs and Outputs
One crucial aspect of using MCP is understanding scenario inputs and outputs. For example, in the image generation scenario, the input is an image prompt. This prompt tells Claude what kind of image to create. In a ClickUp task scenario, inputs include the task title and description. By defining these inputs, Claude knows exactly what information to send when calling the scenarios.
Each scenario also has outputs. For instance, after creating a ClickUp task, the output would be the task URL. This output is then returned to Claude, making it easy to track what actions have been performed. The combination of scenario inputs and outputs streamlines the entire automation process.
Practical Examples of MCP in Action
Now, let’s look at some practical examples of how this all works. After configuring Claude Desktop, I prompted it to create a ClickUp task about AI agents. By using the predefined inputs, Claude populated the task title and description. The response included a ClickUp URL, which I could easily open to verify the task’s creation.
Next, I asked Claude to generate an AI image. I selected the image prompt input and waited for the response. It generated a detailed prompt and created the image, providing me with the URL to access it. This step illustrated how effectively Claude can handle various scenarios using the MCP protocol.
Publishing Content to WordPress
Another exciting capability is publishing content directly to WordPress. After generating the AI image, I instructed Claude to create a WordPress post that included the image. The scenario required inputs like post title, content description, and image URL. I emphasized that the post content must be in HTML format for proper rendering on WordPress.
Once I pressed save, Claude executed the scenario. It successfully created the post and included the image as the featured image. However, I noticed that the heading duplicated in the post. To fix this, I adjusted the scenario input parameters to ensure the HTML content didn’t include the title at the start. After saving the changes, the post was created without any issues.
Access Control and Customization
One of the key features of the MCP is the ability to manage access control. You don’t need to give external systems access to all your “on demand” scenarios. Instead, you can define specific tool access by entering organization IDs or scenario IDs in Claude’s configuration settings. This level of granularity allows you to control which scenarios can be accessed by external agents.
This means you can maintain tighter control over your automation processes while still leveraging the full capabilities of the MCP. I found this feature particularly useful for ensuring that only the necessary scenarios were accessible.
Connecting Other Agents and Systems
Once you’ve grasped the basic concepts of connecting Claude Desktop to Make.com using MCP, you’ll see how it allows you to connect other external agents and assistants with MCP compatibility. This flexibility is invaluable for automating various tasks without the hassle of setting up complex integrations for each tool.
The MCP protocol opens up a world of possibilities. You can stack capabilities from different systems together, enhancing your overall workflow. Whether you’re generating content, managing tasks, or integrating other tools, the MCP makes it easier than ever to streamline these processes.
By using the MCP, I was able to create a comprehensive automation solution that not only saved time but also improved my efficiency. This is a game-changer for anyone looking to leverage AI and automation in their workflows.
In summary, the Model Context Protocol is a powerful feature that simplifies the integration of external agents with Make.com. By understanding how to set up your account, define scenario inputs and outputs, and manage access control, you can unlock a new level of automation. This not only saves time but also enhances the capabilities of your workflows across different platforms.
With the MCP, the future of automation looks bright. Embracing these tools will allow you to achieve more with less effort, making your work more efficient and productive.
