The MCP Server enables AI assistants to trigger and interact with Make automation workflows, exposing them as callable tools.
Connect to Make: Links to your Make account and identifies on-demand scenarios
Parameter Handling: Parses and resolves input parameters, providing meaningful descriptions
Scenario Invocation: Allows AI assistants to run scenarios with required parameters
Output Interpretation: Returns structured JSON output for AI systems to interpret
Maintains Complex Logic: Keeps intricate automation logic in Make while making it accessible to AI
Bidirectional Communication: Facilitates two-way interaction between AI assistants and workflows
Specific capabilities include executing various scenarios like inventory management, dynamic connections testing, and demonstrations of different input types including arrays of collections.
Allows AI assistants to trigger and interact with Make automation workflows, enabling scenarios configured with On-Demand scheduling to be utilized as tools by AI systems
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Make MCP Server run my 'New Customer Onboarding' scenario for john.doe@example.com"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
A modern, cloud-based version of the Make MCP Server is now available. For most use cases, we recommend using this new version.
A Model Context Protocol server that enables Make scenarios to be utilized as tools by AI assistants. This integration allows AI systems to trigger and interact with your Make automation workflows.
The MCP server:
Connects to your Make account and identifies all scenarios configured with "On-Demand" scheduling
Parses and resolves input parameters for each scenario, providing AI assistants with meaningful parameter descriptions
Allows AI assistants to invoke scenarios with appropriate parameters
Returns scenario output as structured JSON, enabling AI assistants to properly interpret the results
Related MCP server: Make MCP Server
Turn your Make scenarios into callable tools for AI assistants
Maintain complex automation logic in Make while exposing functionality to AI systems
Create bidirectional communication between your AI assistants and your existing automation workflows
NodeJS
MCP Client (like Claude Desktop App)
Make API Key with scenarios:read and scenarios:run scopes
To use this server with the Claude Desktop app, add the following configuration to the "mcpServers" section of your claude_desktop_config.json:
{
"mcpServers": {
"make": {
"command": "npx",
"args": ["-y", "@makehq/mcp-server"],
"env": {
"MAKE_API_KEY": "<your-api-key>",
"MAKE_ZONE": "<your-zone>",
"MAKE_TEAM": "<your-team-id>"
}
}
}
}MAKE_API_KEY - You can generate an API key in your Make profile.
MAKE_ZONE - The zone your organization is hosted in (e.g., eu2.make.com).
MAKE_TEAM - You can find the ID in the URL of the Team page.
Tool: List Inventory (Lists items in inventory.)
Inputs: no named parameters.
Tool: Add to Inventory (Add a new item to the inventory.)
Inputs: name (required).
Scenario Inputs All Types
Inputs: number (required), array_of_arrays, array_of_collections, boolean, +6 more.
Scenario Inputs: Array of Collections
Inputs: array.
Dynamic Connections Testing
Inputs: AirtableConnection.
Scenario D - Subscenario
Inputs: text.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 6 tools
The tool names are entirely opaque, consisting only of scenario IDs like 'run_scenario_11422' with no indication of their distinct purposes. From the descriptions, some tools appear to handle inventory operations while others deal with scenario inputs or testing, but the naming provides no disambiguation, making it impossible for an agent to reliably choose between them without guessing.
All tool names follow a perfectly consistent pattern: 'run_scenario_' followed by a numeric ID. While this pattern is uninformative, it is uniformly applied across all six tools, with no deviations in style or structure.
With 6 tools, the count is reasonable and not excessive. However, given the unclear domain inferred from the tool names and descriptions—which suggest a mix of inventory management, scenario testing, and input handling—it's borderline whether this number adequately covers the scope, as the tools seem fragmented rather than cohesive.
Inferred domain includes inventory management and scenario testing, but there are significant gaps. For inventory, only list and add operations are present, missing update and delete. For scenario inputs and testing, the tools are vague and incomplete, lacking clear CRUD or lifecycle coverage, which will likely cause agent failures in complex tasks.
Zero-setup MCP gateway securely connecting AI to your tools with authentication and workflows
Connect your AI assistants to Keboola and expose your data, transformations, SQL queries, ...
Connect your AI agent to 300+ methods and 100+ live tools
Build and run agents and automations across hundreds of apps using natural language.
