n8n-MCP is an independent, MIT-licensed community Model Context Protocol server for assistants that help design and inspect n8n workflows. It is maintained in the czlonkowski/n8n-mcp repository rather than by the n8n product team, and it is distinct from the existing aicoolies record for the n8n automation platform. This page therefore covers the MCP knowledge and workflow-assistance layer, not the base n8n product, its cloud plans, or an official n8n integration owned by n8n GmbH.
The server exposes structured access to n8n node documentation, operations, properties, workflow templates, search, configuration guidance, and multiple levels of node or workflow validation. With optional n8n API credentials it can also support management-oriented workflow operations, while a read-only deployment mode is documented for more governance-sensitive environments. These capabilities can reduce guesswork for an assistant, but they do not make AI-generated workflows safe by default or prove that a proposed node configuration will behave correctly against live services and credentials.
Upstream safety guidance says never to edit production workflows directly with AI: copy the workflow, test in a development environment, export backups, validate changes, and review them before deployment. Self-hosted and Docker installation paths are available, alongside a hosted dashboard that requires account access; because public plan details were not confirmed at write time, this record makes no hosted-price claim. The project also publishes privacy and telemetry controls, so teams should review the opt-out and credential-handling guidance before enabling API-connected tools.
