n8n-MCP
An MCP server that gives AI assistants knowledge of n8n nodes so they can help build n8n workflows.
- Stars
- 23.2k
- Forks
- 3.7k
- Licence
- MIT
- Language
- TypeScript
- Last update
- 4 days ago
- Type
- AI model / agent
What it is
n8n-MCP is a Model Context Protocol server that sits between AI coding assistants (such as Claude Desktop, Claude Code, Cursor, or Windsurf) and n8n, a workflow automation platform. It gives the AI structured access to documentation, properties, and operations for n8n's nodes, plus templates and examples, so it can design and check workflows more accurately. The README says it can be used through a hosted dashboard with a free tier, or self-hosted via npx, Docker, Railway, or a local install.
- Ask an AI assistant to build a draft n8n workflow from a plain-language description
- Search the library of workflow templates (the README cites 2,352) before building from scratch
- Look up node properties and configuration details for specific n8n nodes
- Validate a workflow's configuration before deploying it to n8n
Great for
- People already using n8n who want an AI assistant to draft or check workflows
- Developers using Claude Code, Cursor, Windsurf, or VS Code who want n8n knowledge in their editor
- Users who want to start quickly with the hosted option instead of self-hosting
- Teams that run n8n behind Cloudflare Access or manage an instance-level MCP endpoint
Not ideal for
- People who do not use n8n, since the tool only helps with n8n workflows
- Anyone who wants AI to edit live production workflows without backups; the README explicitly warns against this
- Users who need a fully managed service with no setup, if they prefer not to rely on the hosted tier limits described in the README
How to get it
- 1
Install with a package manager
You'll need: Node.js (version not specified in the provided README excerpt) · An n8n instance or API key if you want the tool to act on your own n8n
Clone the repository
git clone https://github.com/czlonkowski/n8n-mcp - 2
Run it
- Choose the hosted dashboard at dashboard.n8n-mcp.com for the quickest start, or read the Self-Hosting Guide for a local setup
- Sign up and get an API key if using the hosted option
- Add the MCP server to your AI client (for example Claude Code, Cursor, or Windsurf) using the matching setup guide in the docs folder
- Ask your assistant to search templates or build a simple workflow, and review the result before using it
- 3
Get the most out of it
- Use the Claude Project system instructions in the README to guide the assistant's workflow process
- Install the n8n-skills companion repository for specialized workflow-building skills
- Set N8N_CF_CLIENT_ID and N8N_CF_CLIENT_SECRET if your n8n instance is behind Cloudflare Access
- Configure N8N_MCP_ACCESS_TOKEN to use the instance-level MCP tools such as n8n_manage_agents
Downloads
Strengths
- Covers a large catalog of n8n nodes, including community nodes, according to the README
- Works with several popular AI coding tools and has setup guides for each
- Open source under the MIT license with a large community, as shown by its star count
Trade-offs
- AI output can be unpredictable, and the README itself recommends testing in a copy of the workflow first
- Coverage figures in the README (such as operations at 66.5%) mean some node actions are not fully documented
- Self-hosting and n8n-instance connections require API keys and some configuration
How it compares
| Project | Stars | Type | Level | Updated | Best for |
|---|---|---|---|---|---|
| n8n-mcp (this) | 23.2k | AI model / agent | intermediate | 4 days ago | People already using n8n who want an AI assistant to draft or check workflows |
| context7 | 64.1k | AI model / agent | beginner | 4 days ago | Developers who use AI coding agents such as Claude, Cursor, or VS Code for libraries that change often |
| apify-mcp-server | 14.3k | AI model / agent | beginner | 4 days ago | Developers who want an AI agent to pull data from websites without writing scrapers themselves |
| fastapi_mcp | 12k | Library | intermediate | 12 months ago | Python developers who already have a FastAPI app and want AI assistants to use it |
| codebase-memory-mcp | 48k | Dev tool | intermediate | 4 days ago | Developers who use an MCP-capable AI coding agent on a large codebase |
| github-mcp-server | 34.2k | Dev tool | intermediate | 4 days ago | Developers who already use an MCP-compatible AI tool such as VS Code with Copilot, Claude Desktop, or Cursor |
Stats synced from GitHub. The assessment above was drafted with AI from the project's README on 2026-10-29 and may miss recent changes — check the README before relying on a command. MIT licence: free to use, modify and distribute, including commercially, as long as you keep the copyright notice.