Delv
Official (Vendor)Active· 9d4.3by Make (Integromat)

Make MCP

Turns Make automation scenarios into callable tools for AI assistants, bridging 1,000+ apps through Make workflows.

A
Safety & Trust

Delv Safety Grade: A

Score 84/100 · assessed 2026-04-18

Maintainer95
Permissions75
Supply chain85
Transparency80
Incidents100

Make (formerly Integromat) is a well-established automation platform owned by Celonis, serving enterprise customers globally. This official MCP server allows Claude to trigger Make scenarios, effectively bridging 1,000+ integrated apps. The maintainer score is excellent given Make's market position and professional operations. Permissions are moderately scoped: the server executes pre-built Make workflows via API, which themselves may perform writes across multiple services (messaging, databases, file storage) depending on scenario configuration. Supply chain is solid with npm distribution and npx installation, though dependency pinning details weren't verified. Transparency is good with open-source code on GitHub, though documentation could be more comprehensive. No known security incidents. The main consideration is that workflow execution scope depends entirely on what the user has configured in Make, creating variable permission boundaries. Users should audit their Make scenarios before exposing them to AI assistants.

Green flags

  • Official vendor (Make/Celonis) with enterprise-grade platform
  • Distributed via npm with standard npx installation
  • Open source repository with clear GitHub presence
  • Leverages existing Make platform security and audit capabilities
  • No known security incidents or CVEs

Red flags

  • Workflow permissions depend on user's Make scenario config (variable scope)
  • API key grants access to all user's Make scenarios without granular control
  • Executing arbitrary workflows could trigger unintended multi-app actions

Permissions requested

Outbound networkAccess secretsSend messagesDB writeWrite filesExternal LLM call
Assessed by Delv Editorial using public metadata. Grades are advisory and update as the ecosystem changes. They do not replace your own review of permissions and code before granting an agent access to sensitive systems.

Install

npx -y @makehq/mcp-server
Env vars needed: MAKE_API_KEY

Review

Make MCP turns your Make scenarios into callable tools that Claude or Cursor can trigger directly. If you've already invested in Make (formerly Integromat) workflows, this is the fastest way to let an AI assistant fire them off without writing custom API wrappers. You point the server at your Make account via an API key, and it exposes your scenarios as MCP tools. Claude can then trigger a workflow to, say, add a Notion page, send a Slack message, and update a Google Sheet in one go. The core value is access to Make's 1,000+ app connectors without building individual MCP servers for each. I'd reach for this if I already have a library of Make scenarios doing real work. For example, a "new lead" workflow that writes to Airtable, posts to a CRM, and emails the sales team becomes a single tool Claude can invoke. The server doesn't build scenarios for you, it just makes existing ones callable. That's a narrower use case than it sounds, but for teams already on Make, it's a clean bridge. Quirks: you're limited by Make's execution model. Scenarios run asynchronously, so you won't get instant feedback in the chat. If a workflow takes 30 seconds to complete, Claude won't know the result unless you poll or set up a webhook. The server also assumes you've already designed the scenario with sensible input parameters. If your Make workflow expects a JSON blob with 12 fields, Claude will need to construct that correctly, which can get fiddly. Who shouldn't bother: if you don't already use Make, this isn't a reason to start. Building a Make scenario just to wrap it in MCP is overkill when you could write a direct integration. This is for teams with existing Make infrastructure who want to add conversational triggers. It's also overkill if you only need one or two simple automations. A custom MCP server or a direct API call will be faster to set up and debug.
Verdict

Install this if you already run Make scenarios and want Claude to trigger them. Skip it if you're starting from scratch or only need a handful of simple automations. It's a bridge, not a builder.

Good at

  • Instant access to 1,000+ app connectors if you already use Make.
  • No need to write individual MCP servers for each integration.
  • Official vendor support means it's maintained alongside Make's own platform.
  • Works with existing scenarios, so no need to rebuild your automation logic.

Watch out

  • Only useful if you already have Make scenarios built and running.
  • Asynchronous execution means no immediate feedback in the chat.
  • Requires careful scenario design with clear input parameters for Claude to use.
  • Overkill for simple one-off automations where a direct API call would suffice.

Use cases

  • multi-app automation
  • workflow triggers
  • data sync
  • no-code orchestration

Getting started

1. Generate a Make API key from your Make account settings (under API section). 2. Run `npx -y @makehq/mcp-server` and follow the prompts to authenticate with your API key. 3. Add the server to your Claude Desktop config (usually at `~/Library/Application Support/Claude/claude_desktop_config.json`) with the command and env var `MAKE_API_KEY`. 4. Restart Claude Desktop and check the MCP tools list to confirm your Make scenarios appear as callable tools. 5. Watch out: scenarios run asynchronously, so Claude won't see immediate results unless you build a polling mechanism or webhook into your workflow.

Works with

Claude DesktopClaude CodeCursor

Similar MCPs