Hjarni for n8n

The AI-native knowledge base for n8n

n8n automates the work around your AI. Hjarni is the knowledge those workflows read and write, so an agent step acts on your real notes instead of a static prompt.

Get started free See how it works

Free to start. No credit card required.

What this unlocks

Workflows that actually use your context

Agents that act on real context

An n8n AI Agent can call Hjarni mid-workflow to read the notes that matter for the task, instead of working from whatever a node hard-codes into the prompt.

Write outcomes back automatically

Hjarni's MCP tools include create and update. A workflow can log a summary, a decision, or a result to a note at the end of a run, no human step required.

Keep automations and humans in sync

The same Hjarni notes your team reads in ChatGPT and Claude are the ones your n8n workflows read and update, so the automation does not drift from the source of truth.

Reusable memory across workflows

Point multiple workflows at the same knowledge base. Conventions and reference notes live in one place rather than being copied into each automation.

Setup

Connect n8n in about two minutes

1
Sign up for Hjarni. Decide whether to authenticate with OAuth2 or a personal API token.
2
In an n8n AI Agent workflow, add the MCP Client Tool sub-node.
3
Set the server endpoint to https://hjarni.com/mcp using the HTTP Streamable transport.
4
Add the credential: OAuth2, or a Bearer header carrying your Hjarni API token.
5
Choose which Hjarni tools to expose to the agent, then run the workflow and check that it can list notes.

MCP endpoint for the MCP Client Tool node

https://hjarni.com/mcp

In an AI Agent workflow, add the MCP Client Tool sub-node and set the endpoint to the URL above (HTTP Streamable transport). Choose OAuth2 or a Bearer credential with a personal API token from Hjarni settings. Field labels in the node change over time, so follow n8n's MCP Client Tool docs for the current shape.

Why an automation platform benefits from a writable knowledge base

n8n is good at moving data between systems and running agent steps. What it usually lacks is a durable place for the facts those steps depend on: the conventions, the reference material, the decisions that should outlive a single run.

Hjarni is that place. Plain Markdown notes you own, readable and writable over an MCP server, so a workflow can both pull context in and push results back out as structured notes.

n8n's MCP Client Tool node connects to remote servers with OAuth2 or Bearer auth, so Hjarni slots in as one endpoint. The notes your workflows touch are the same ones your people read in their AI clients.

Common questions

Questions before you connect n8n

How do I connect n8n to Hjarni?

Add the MCP Client Tool sub-node to an AI Agent workflow and point it at https://hjarni.com/mcp using the HTTP Streamable transport.

What authentication does the MCP Client Tool node support?

OAuth2 and Bearer, among others. Use OAuth2, or a Bearer credential carrying a Hjarni personal API token.

Can an n8n workflow write notes, not just read them?

Yes. Hjarni's MCP tools include create and update, so a workflow can log a summary, decision, or result back to a note at the end of a run.

Is this the same as n8n's MCP Server Trigger?

No. The MCP Server Trigger turns an n8n workflow into a server other clients call. To consume Hjarni, n8n is the client, so you use the MCP Client Tool node.

Will the node's fields match the docs exactly?

n8n's node labels change over time. Follow the official MCP Client Tool docs for the current transport field. The endpoint and auth choices above stay the same.

Give n8n a memory

The session starts where the last one ended. Write notes once. n8n reads them across every conversation.

Get started free See pricing

Give your AI a memory

Write once. You both remember.