Creativity · MCP — server

MCP n8n Server

The MCP n8n Server is a community-maintained Model Context Protocol server that wraps the n8n REST API. It lets an LLM client browse workflows, trigger runs, fetch execution logs, and inspect node output — useful both for running existing automations and for letting an agent test flows it is iteratively building.

MCP facts

Kind: server
Ecosystem: anthropic-mcp
Language: TypeScript / Node.js
Transports: stdio

Capabilities

Tools: list_workflows, run_workflow, get_execution, list_executions, activate_workflow
Resources: n8n://workflow/{id} URIs
Auth: n8n API key with workflow:run and workflow:read scopes

Install

npx -y @modelcontextprotocol/server-n8n

Configuration

{
  "mcpServers": {
    "n8n": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-n8n"],
      "env": {
        "N8N_URL": "https://n8n.example.com",
        "N8N_API_KEY": "..."
      }
    }
  }
}

Frequently asked questions

Is this server official?

It is community-maintained. n8n also ships its own 'AI Agent' and MCP node that let workflows consume MCP servers — the inverse direction.

Should I self-host or use n8n Cloud?

Both work. Self-hosted gives more control over secrets and webhook URLs; Cloud is faster to set up. API surface is the same.

Does triggering a workflow cost credits?

On n8n Cloud yes — executions count against the plan quota. Self-hosted instances are only limited by your server.

Sources

n8n API docs — accessed 2026-04-20
MCP community servers — accessed 2026-04-20