An MCP that makes Claude read the n8n manual so you don't have to

What it does

n8n-MCP is a Model Context Protocol server that translates n8n’s sprawling automation catalog into something AI assistants can actually reason about. It exposes schemas, documentation, and real-world examples for 1,851 nodes—822 core plus 1,029 community integrations—letting models like Claude search, configure, and validate workflows instead of guessing JSON shapes. A hosted dashboard exists for those who want instant access without managing infrastructure.

The interesting bit

The depth is almost obsessive: 99% property coverage, 87% documentation coverage, and 2,352 ranked workflow templates with near-complete AI metadata. Rather than letting the model freestyle, the server enforces a validation pipeline—quick required-field checks, then full runtime simulation, then workflow-wide connection and expression audits—before anything touches your n8n instance. It is essentially a pedantic build engineer compressed into an LLM context window.

Key highlights

• Structured access to 1,851 n8n nodes (822 core, 1,029 community) with 99% property schema coverage
• 2,352 workflow templates and 156 real-world ranked configurations drawn from popular setups
• Staged validation: node-level checks, workflow connection graphs, and expression validation
• Hosted tier at dashboard.n8n-mcp.com with daily free limits; self-hosting supported via Docker or Railway
• Plugs into Claude Code, Cursor, Windsurf, VS Code Copilot, and Codex

Caveats

• Node operations coverage is 63.6%, so some actions may be absent from the AI’s context
• Real-world example availability varies by node popularity; obscure integrations can come up empty
• The maintainer explicitly warns never to let AI edit production workflows directly—copy, test in dev, and backup first

Verdict

Grab it if you orchestrate complex n8n flows and want an AI assistant that actually knows the node catalog instead of improvising. If your automation needs stop at a single webhook, it is probably overkill.