n8n

The MCP Client Tool node, on an AI Agent node.

Kind MCP client

Transport MCP Client Tool node, HTTP Streamable.

Config The workflow canvas

Setup

Add an AI Agent node to the workflow.
Attach an MCP Client Tool node to its tool input.
Set the endpoint below and choose HTTP Streamable as the transport.
Create a Header Auth credential with name x-api-key and your key as the value.

Endpoint:    https://api.unzoi.com/mcp
Transport:   HTTP Streamable
Auth:        Header Auth
  Name:      x-api-key
  Value:     nai_your_key_here
Tools:       All  (or select the ones this workflow needs)

What differs here

The tool-selection field on the node is worth using: a workflow that only ever needs clustered stories should expose only list_stories, so the model cannot spend quota elsewhere.
The node retrieves the tool list when the workflow runs, so a key that is wrong shows up as a failing node rather than a missing tool.
n8n also has an MCP Server Trigger for the opposite direction — exposing your own workflows as MCP tools. It is not what you want here.

Next

Get a key at console.unzoi.com, if you have not.
The tool reference — what the model is reading when it chooses between the six tools.
Stories for context windows — the one habit that most changes how well an agent works against this API.
Limits — an agent that loops can spend a monthly quota faster than you expect.
n8n's own MCP documentation — authoritative for anything about the client itself, including config paths, which move.

Endpoint, for copying: https://api.unzoi.com/mcp