Hjarni

What I Actually Use Claude's Note Access For

A few months ago I connected Claude to my notes via MCP. People keep asking what I actually use it for. Here's the honest answer.

Reusable context

I have a note with my stack, my architecture decisions, my constraints. Claude reads it automatically. Every new conversation starts informed. No more pasting the same background over and over.

That alone was worth setting up.

Decisions log

I keep a note with decisions and the reasoning behind them. Not just "chose SQLite" but why. Claude works within constraints I've already thought through instead of reopening closed questions.

Writing in my own voice

My copywriting rules live in a note. Short sentences. No jargon. Opinionated. Claude writes copy that sounds like me on the first try. The editing time dropped significantly.

What didn't stick

Meeting notes. I tried it, stopped after two weeks. MCP memory works best for knowledge that's stable — your stack, your principles, your decisions. Not things that change every week.

The pattern

Write once. Use everywhere. The notes that pay off are the ones you'd otherwise re-explain. Write them down once in a place Claude can read. Stop repeating yourself.

That's what Hjarni is for. Simple knowledge base, built-in MCP server. Try it free.