As LLMs get better at drafting plans, reviews and emails, waving their output through gets easier. That is how you end up shipping something you never actually read (and in doing so, annoying your colleagues or readers).
To avoid that, I got into the habit of asking a model to write a file to my desktop for a proper edit and review. That works alright, but reading raw markdown is a chore, no WYSIWYG editor I found looked good, and every question or task I wanted to delegate meant copying text back into a terminal.
Over coffee, I worked with Claude to create Sidenote, a piece of personal software in the form of a minimalist markdown editor mac app that connects to a Claude Code session.
It allows me to comment on a line to question it or ask for a change, and have the session reply in the comment thread. A pleasant side effect is that multiple comment threads can run at once, each pinned to the passage it is about. This contrasts with my experience of a single-threaded chat session, which keeps the same exchanges in one column: the model has no trouble following it, but it leaves the human scrolling back to find where something was discussed.
Versions of this collaborative UX are scattered across the Claude apps — comments on a plan file, threads on a published artifact, tracked changes in Word — but none of them start from an editor I actually want to write in. That is the part I wanted: a good editor first, that happens to collaborate.

I’m writing this in Sidenote right now!
When updating my site with this project, I realised the page needed some copy, and I asked the model to open Sidenote so I could do a first draft. When I’m done I’ll leave a comment, and the session pushes the change – a neat hand-off between a model and a human.
Apple silicon, macOS 13 or later.
brew install --cask benedictshegog/sidenote/sidenote