macOS xterm-256color zsh 112 views

I’ve been working with LLMs in text files instead of chat interfaces, and it unlocked something useful: inline commenting.

Instead of restating context or copying parts of the AI’s response into your next message, you can just comment directly within the assistant’s answer using simple markers like “>>”. Then prompt with “see comments inline.”

It treats the conversation like a living document—similar to code review or Google Docs comments—and makes iteration way faster and clearer.