# n8n

An MCP server that hands your AI assistant n8n node docs, properties and operations, plus a library of ready workflow templates.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI
- Author: czlonkowski
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/czlonkowski/n8n-mcp
- Updated: 2026-10-29
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/mcp/n8n

## What it does

It answers the question every chat-driven n8n session hits: what parameters does this node take, and which operations can it run? Your assistant gets structured data for 3,143 nodes (841 core, 2,302 community) with property schemas and operation descriptions. It also reads 2,352 workflow templates and 156 ranked configurations taken from popular ones.

The tools cover node and template lookup, config checks through `validate_node` in `minimal` and `full` mode, and n8n agents over instance-level MCP. That last part takes `N8N_MCP_ACCESS_TOKEN` from the n8n instance settings, a secret separate from `N8N_API_KEY`. Instances behind Cloudflare Access want `N8N_CF_CLIENT_ID` and `N8N_CF_CLIENT_SECRET`.

Operation coverage sits at 66.5%, so some actions your assistant will fill in from guesswork. The author's warning is blunt: keep AI out of production workflows, work on a copy and test in development first. A hosted option runs on `dashboard.n8n-mcp.com` with 100 tool calls a day on the free tier; self-hosting goes through `npx`, Docker or Railway.

## Requirements
- Needs: Docker
- Works with: n8n

## Install

### Claude Code (native platform)

**Install**
- The MCP registry has no launch command for it: see the repository README

## When to use it
- look up a node's parameters instead of guessing them
- start a workflow from an existing template
- validate node config before deploying
- check which operations a community node supports
- wire up n8n AI tools and read their schemas

## Example prompts
- Find a workflow template for syncing data and adapt it to my instance.
- Run `validate_node` in `full` mode on this workflow before I publish it.
- List the operations and parameters of the `HTTP Request` node.

## Similar
- [n8n-subworkflows-official](https://affpapa.org/en/nevibecode/skills/n8n-subworkflows-official.md) — skill · 2300 installs
- [n8n-multi-instance](https://affpapa.org/en/nevibecode/skills/n8n-multi-instance.md) — skill · 2300 installs
- [using-n8n-mcp-skills](https://affpapa.org/en/nevibecode/skills/using-n8n-mcp-skills.md) — skill · 2100 installs
- [cli-anything-n8n](https://affpapa.org/en/nevibecode/skills/cli-anything-n8n.md) — skill · 593 installs
- [n8n-workflow-automation](https://affpapa.org/en/nevibecode/skills/n8n-workflow-automation.md) — skill · 121 installs
- [n8n-mcp-synta-codex](https://affpapa.org/en/nevibecode/plugins/n8n-mcp-synta-codex.md) — plugin · ★ 1596 repo stars

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
